Sign up

Monetary transmission mechanism and behaviour of asset prices: the case of Croatia.
Subject:
Monetary policy
Interest rates
Bank loans
Going public (Securities)
Economic indicators
Construction industry
Entrepreneurship
Authors:
Ivanov, Marijana
Lovrinovic, Ivan
Pub Date:
03/01/2008
Publication:
Name: Review of Business Research Publisher: International Academy of Business and Economics Audience: Academic Format: Magazine/Journal Subject: Business, international Copyright: COPYRIGHT 2008 International Academy of Business and Economics ISSN: 1546-2609
Issue:
Date: March, 2008 Source Volume: 8 Source Issue: 2
Topic:
Event Code: 900 Government expenditures Computer Subject: Company public offering
Product:
Product Code: E561000 Interest Rates; 8910000 Engineering & Architectural Svcs; 1500000 Construction NAICS Code: 54133 Engineering Services; 23 Construction SIC Code: 1521 Single-family housing construction; 1522 Residential construction, not elsewhere classified; 1531 Operative builders; 1541 Industrial buildings and warehouses; 1542 Nonresidential construction, not elsewhere classified; 1623 Water, sewer, and utility lines; 1629 Heavy construction, not elsewhere classified

Accession Number:
182406762
Full Text:
ABSTRACT

Asset price channels have gained significant importance in the activity of the monetary transmission mechanism. Changes in the aggregate demand are determined by the movement of interest rates, but also by the movement of other prices such as stock prices, foreign exchange rates, prices of housing and land. As a result of the abundance of global liquidity, countries of the so-called emerging markets have recorded high rates of monetary growth as well as strong credit expansion for several years now. As a result of the effect of domestic and imported factors on the increase in liquidity, credit expansion acts as a trigger of money supply growth, and generator of increased spending and growing possibilities of savings and investments as a result of the wealth effect. Credit and monetary expansion increases demand for financial and non-financial forms of assets, which stimulates rises in asset prices. In the cause-effect relationship, the increase in the market value of assets used as collateral enables businesses or households to borrow more and results in the periods of continuous expansion of bank credits over the years. In circumstances where the goal of most central banks is oriented only toward maintaining the stability of prices defined in terms of consumer price indexes, the rise in asset prices (shares, commercial real estate, residential real estate etc.) can escape regulation of the central bank and institutions of prudential control for a long time, and can transmit effects on the rise in other prices, i.e. labour, utilities, etc. In such circumstances the credit channel of monetary transmission obtains special significance in the analysis of factors of destabilisation of the real economy.

Keywords: Monetary Transmission Mechanism, Asset Prices, Behavioural Finance

1. INTRODUCTION

Global flows of capital modify the significance of particular channels of the monetary transmission mechanism and highlight new dimensions of the liquidity effect and the effect of real money. In particular, asset price channels have gained significant importance in the activity of the monetary transmission mechanism in the past twenty years. As a result, liquidity transmission through the financial system and the economy has somewhat changed, and so have initial and feedback money effects on the movement of consumer prices and asset prices. Changes in the aggregate demand are determined by the movement of interest rates (price of money--price of debt), but also by the movement of other prices such as stock prices, foreign exchange rates, prices of housing and land. Therefore, the asset price movement depends on fundamental reasons which influence their real value, but also on the financial system liquidity arising not only from domestic expansionary monetary policy but also from the import of liquidity abundance from the global market.

With stronger liberalisation of capital flows, significant foreign exchange inflows/outflows increase the importance of monetary effects of foreign exchange transactions as a flow of creation/destruction of primary money and money supply. At the same time, the practice of managed exchange rate floating decreases the opportunities of central banks to establish control over money supply growth which is generated by strong credit activity of banks. Credit expansion triggers money supply growth, and generates growth in spending and appreciation of asset prices through the wealth effects.

Credit and monetary expansion increases demand for financial and non-financial forms of assets, which in turn stimulates increases in asset prices. In the cause-effect relationship, an increase in the market value of assets used as collateral enables more borrowing by businesses and households and results in periods of continuous growth of bank credits over the years. The credit channel of monetary transmission operates by means of the balance sheet channel, which deepens the relationship between asset price movements and movements in the real sphere of the economy. Companies whose shares quote better on the stock exchange and which have a better financial standing gain an opportunity to borrow even more and under more favourable conditions (lower effective interest rates). Citizens whose real estate gains value can choose between new borrowing possibilities for the purchase of new or bigger real estate, for the acquisition of a vessel etc. At the same time, due to the growth of asset value, the same real estate serves as collateral necessary for obtaining a higher amount of loans or more loans.

2. LITERATURE REVIEW

Generally speaking, the monetary transmission mechanism represents the mechanism of transmission of monetary impulses on the financial system and the real economy. It is manifested through several types of transmission channels: interest rate channel, asset price channel, bank lending channel, balance sheet channel and household liquidity channel.

In accordance with the traditional understanding of the concept, the liquidity effect occurs as a result of monetary transmission through the interest rate channel. According to Keynesian interpretations (see for example Dornbusch and Fischer, 1987) liquidity growth (a more expansionary monetary policy) decreases nominal interest rates and enables more investments and more household and government spending. This leads to a two-way stimulating effect between money supply growth and output growth. Money supply growth leads to increases in output and greater output stimulates further growth of the money supply and demand for liquidity. In the circumstances of an expansionary monetary policy and expectations of further weakening of monetary constraints, the fall in short-term interest rates is reflected on the fall in long-term interest rates. With that respect, a fall in real interest rates coupled with falling nominal interest rates is very important for the practical efficiency of the interest rate channel and liquidity effect. At the same time, the substitution effect and income effect occur. For instance, in the case of falling interest rates, the substitution effect favours current spending in relation to savings. The income effect influences increases in available income and spending possibilities. Monetarists (Friedman and Schwartz, 1963) stress the feedback effect through which increased demand for money subsequently stimulates the growth of nominal interest rates. As a result, the Fisher effect occurs in the long run related to the growth of inflation expectations and the effect of real money.

Speedy development of financial systems, the phenomenon of the so-called financial innovations (derivatives, securitisation ...) and a growing degree of liberalisation of international financial transactions increase the importance of the asset price channel and the activity of asymmetric information in the monetary transmission mechanism. As a result, the transmission of liquidity through the financial system and the economy partially changes, as well as the initial and feedback monetary effects on the movement of consumer prices and asset prices.

For the past twenty years, apart from interest rate movements, changes in the aggregate demand and income effects have been increasingly determined by stock price movements, housing price movements, land price movements as well as the price of foreign money in terms of domestic currency. All this is in accordance with monetarist views according to which changes in monetary policy simultaneously influence prices in different markets of financial assets and durable goods, and particularly in the stock market and real estate market. Therefore, changes in the prices of these goods can cause major and considerable effects on output and employment through spending.

According to Mishkin (2004) there are three types of asset channels: foreign exchange channel, Tobin's Q ratio channel and wealth channel. In the case of expansionary monetary policy and fall in interest rates, the foreign exchange channel reflects depreciation effects on net export growth, and through effects on the prices of imported goods, it affects the formation of the aggregate level of prices in the national economy. Tobin's Q ratio channel includes effects of increased money supply on the public demand for different forms of assets and subsequent effects of asset price changes on economic movements. According to Tobin's research (1969), high stock prices allow businesses smaller costs of financing investments through issuing of new shares. Similar increases in real estate prices, which increase their market value in relation to construction costs, lead to bigger investments in the construction sector, which has a positive effect on national income. The growth of the market value of shares and real estate increases the wealth of their owners in monetary terms. The growth of wealth stimulates the purchase of new shares or real estate, and generally fosters increased household spending especially in the segment of increased expenses for the acquisition of durable goods.

According to research conducted by Lusardi (1996) and Souleles (1999) "changes in housing wealth might have a larger effect on consumption than changes in stock market wealth". The reason for that is linked with the fact that purchasing housing real estate is characteristic for a larger number of people than the acquisition of shares, bonds or other financial forms of assets. In the analysis of the relationship between monetary transmission and housing price movement, Mishkin (2007) stresses the importance of the distributional effect on the basis of which the effect of growing wealth on the growth of spending depends on who the buyer of the real estate is, or whether the purchase of real estate is connected with mere subsistence and housing needs or with the speculation on the potential future growth of real estate prices.

The interaction of the interest rate channel and asset channel reflects the presence of asymmetric information between borrowers and lenders, which affects the imperfection of financial markets and results in unexpected shocks in the economy (Lovrinovic and Benazic, 2004). The effect of asymmetric information is manifested in the form of several derived channels of the monetary transmission mechanism. The first among them is the bank lending channel derived from the interest rate channel. In the circumstances of an expansionary monetary policy, bank liquidity grows and so does the bank credit supply in the market. According to Azali research (1998), increased credit supply and consequent effects in the direction of increased money supply stimulate the growth of output and characterise the entry phase of the business cycle by means of a two-way cause-effect relationship.

Commercial banks choose clients to whom they will grant credit depending on risk assessment, and therefore banks are in the position to direct monetary credit funds, which can result in asymmetries in the economy. Ivanov and Cavrak (2005) examine bank preferences as a variable of the credit supply function which leads to imbalances between credit amounts granted to different sectors. Thus for example, the credit channel can strongly affect the growth of household spending, and to a lesser extent it can enable the growth of business investments. Furthermore, banks can prefer crediting big business, and deny money to small and medium size enterprises.

Asymmetry of information between borrowers and lenders is especially manifested in the effect of the balance sheet channel. Expansionary monetary policy and subsequent fall in interest rates increase household spending and enable the growth of revenues for businesses. With bigger net cash flows the financial position of a business as a borrower improves, due to which external finance premium decreases. All of the above enables borrowing even by those businesses for which earlier external finance premium was very high. Furthermore, the fall in interest rates affects the growth of market stock prices and real estate prices, which in credit business represents an increase in the value of assets which serve as collateral in borrowing. The increased value of assets used as collateral enables easier financing of investments, with the result that more and more projects enter the circle of cost-effective investments, and the circle of good debtors expands to include more and more clients. All this, coupled with the assumption that the company owner's interest is to reduce the risky behaviour of the company management (in order to maintain the high share price), increases the interest of banks in crediting the company at more favourable interest rates and under other favourable conditions. However, since borrowers and lenders have asymmetric information, borrowers' balance sheets may depict too optimistic a state due to which assessments of client creditworthiness may be too optimistic. Historical experience proves that such overoptimistic assessments of credit risk were the main factor in declaring bank illiquidity, and in more drastic circumstances, they led to bank insolvency, bankruptcy and rise of the bank crisis.

The interdependence of micro and macro aspects of liquidity is linked to another segment of activity of the monetary transmission mechanism in the form of household liquidity channel. In conditions when monetary constraints are getting more lax, readiness of households to increase spending grows. Reasons for that include easy access to credit, lower debt service costs, increase in the market value of real estate and stocks, high liquidity of stocks in the market, and expectations of future positive movements. In conditions of household net wealth growth and household income growth (at least in the nominal sense), financial disruptions are less likely to occur. Since the financial position of individuals/households is safer, and the possibility of financial difficulties is smaller, the psychological effects of abstaining from a desired purchase decrease and so does their interest in savings for the rainy day. At the same time, expenses for different goods and services increase as well as interest in the purchase of durable goods and real estate.

According to Gabe De Bondt's approach (2000), the effect of monetary transmission is also determined by the so-called channel of expectations and insecurity related to anticipated or non-anticipated measures of monetary policy, due to which the described channels do not always operate in the direction mentioned and due to which monetary policy measures do not lead to desired effects.

In the circumstances of monetary expansion, the expected result of effects of all desired transmission mechanism channels is bigger output. The described mechanism of the so-called financial accelerator is related to the role of money supply as the generator of economic expansion. However, its effects operate in the opposite direction as well--the downward trend of the economy when the fall in the money supply resulting from the panic in the financial sector stimulates the effect of deepening recession.

The unfavourable effect of the financial accelerator also occurs because of the feedback of the effect of real money, which in the well-known IS-LM model reflects the growth of nominal interest rates and the return of the LM curve to the left toward lower income levels. The fall of the real effect of money occurs due to inflation effects generated by the income effect, increased spending and increases in prices of goods and services, as well as the growth of nominal wages in the labour market or other causes of inflation. The intensity of the adverse effect of real money in relation to the intensity of the positive liquidity effect depends on several factors. Primarily these factors include the intensity and speed of inflation expectations of the public and the negotiating power (ability) of participants of the labour market and market of goods and services to build the inflation premium into the existing and future prices.

Conflicting attitudes of neokeynesian and neoclassical (monetarist) theory on the effectiveness or neutrality of monetary policy in the short (and long) run, ground their explanations of the importance of the effect of real money on the rational behaviour of participants of all three markets, and on the sustainability of the thesis of stickiness in prices in the labour market and in the market of goods and services. Increasing the degree of rational behaviour of all market participants and increasing the negotiating power of the agents of supply of trading objects (labour, goods/services, and money/capital) raises the power of the effect of real money due to inflation effects. The development of an inflation spiral can be determined by the combination of various factors: increased liquidity of the economy and money supply growth, increased demand and spending of all sectors, increased labour costs, increased input purchase costs, rise in the prices of energy products and essential food products, increased costs of borrowing, public expectations, asset price movements, uncertainty of future asset price movements etc.

3. LIQUIDITY OF THE ASSET MARKET

As a result of the abundance of (global) liquidity, countries of the so-called emerging markets have recorded high rates of monetary growth as well as strong credit expansion for several years now. As a result of the effect of domestic and imported factors on the increase in liquidity, credit expansion acts as a trigger of money supply growth, and generator of increased spending and growing possibilities of savings and investments as a result of the wealth effect. Credit and monetary expansion increases demand for financial and non-financial forms of assets, which stimulates rises in asset prices.

In the cause-effect relationship, the increase in the market value of assets used as collateral enables businesses or households to borrow more and results in the periods of continuous expansion of bank credits over the years. The credit channel of monetary transmission acts here through the balance sheet channel, which deepens the link between asset price movements and movements in the real sphere of the economy. Companies whose shares quote better on the stock exchange and which enjoy a strong financial standing obtain the opportunity for additional and more favourable conditions of borrowing (lower effective interest rates). Citizens whose real estate gains more value can choose between different new possibilities of borrowing--for the purchase of new or bigger real estate, for the purchase of a vessel etc.

At the same time, due to the growth of value the same assets are used as collateral for bigger amounts of credit (conversion of credits approved earlier into credits of higher amounts, granting new credits, etc.).

There are a number of reasons which stimulate commercial banks to continually provide new sources of liquidity to be able to finance increases in assets and further market growth. After exhausting domestic sources of funds (primarily citizens' deposits) banks raise funds by borrowing abroad, or by securitisation and other techniques. All of the above is especially affected by the internalisation of banking business, global processes of acquisitions and mergers in the financial industry, growing competition in the market, the need to find new fertile soil for the growth of profit and continuation of further liberalisation of capital flows.

The result of additional effects of obtained liquidity includes long-term effects of the liquidity effect on the growth of asset prices. During the 1980s, major expansion of credits to households and businesses brought about the boom in asset prices in the markets of USA, Great Britain, Japan, Scandinavia, the Netherlands and elsewhere. During the 1990s, liberalisation of capital flows facilitated access to capital by less developed countries of the so-called emerging markets, which led to significant growth of asset prices, amount of credits and spending in the economies which emerged as net recipients of global capital. (In that mention, we should not forget cases in which due to flaws in regulation and inability to defend the financial system from shocks, some form of financial crisis occurred which led to price corrections and significant fall in the value of assets as a result of growing instability in the banking sector).

With the abundance of liquidity, the effects of magic thinking and irrational optimism (that positive trends will last forever), as well as other forms of behaviourist determinants of the movements of share and real estate prices, the period of the upward asset price movement trend is prolonged. As a result asset prices are likely to stay misvalued for a long time and retained at levels above and far above the fundamental value. The efficiency of the financial market thereby becomes significantly reduced and puts under question the accuracy of analyses of future price movements based on assessments of the fundamental value, i.e. fundamental factors of the direction of asset prices.

In accordance with the controversial efficient market hypothesis, asset prices remain at their fundamental level at all times and reflect all the information related to past and future events, due to which there is no possibility of arbitration. Share prices reflect all publicly available relevant information related to the business activity of a company such as quarterly and annual reports, announcement of dividends, financial, business and political news on the changes of management, new product lines etc. According to Fama (1969) in the extreme form of the so-called high efficiency of the share market, prices reflect all public and private information relevant for the business activity of a company, including secret insider information. Similarly, real estate prices reflect all historic and current relevant information available to the general public. Since information is at the same time known to numerous market participants, in conditions of an efficient market no one is in the position to make extra profit arising from the attempt to seek undervalued or overvalued shares--there is simply no free lunch. New information cannot be foreseen and the movement of prices is unpredictable and reflects a random walk in which the level of prices can go up or down. If possibly short-term opportunities for arbitration appear due to the domination of rational and informed investors who quickly buy undervalued assets and sell overvalued assets according to new information, asset prices are corrected quickly and return to their fundamental value, due to which most of the time assets are valued realistically.

In refutation of the efficient market hypothesis, research conducted by behavioural finance advocates shows that asset prices greatly reflect both the activity of non-fundamental factors which stimulate the processes of overvaluation or undervaluation of particular assets. High volatility of asset prices and misevaluation of assets caused by non-fundamental reasons is often the result of weak regulatory practice, non-transparent legal frameworks and non-transparent accounting system. Due to the influence of irrational behaviour of part of participants in the financial and real estate market, long-term periods may occur in which levels of asset prices are misvalued, and which subsequently trigger different turbulences in the financial market, but also business cycle shifts in particular segments of the economy. In conditions of high liquidity of the financial system, share liquidity and real estate liquidity grow which together with high trading volume and high demand stimulate the increase in prices above levels which correspond to the real value. Representatives of behaviourist finance mention a number of psychological factors which trigger irrational behaviour of the so-called noise traders and which cause overvaluation of assets.

In circumstances where the goal of most central banks is oriented only toward maintaining the stability of prices defined in terms of consumer price indexes, the rise in asset prices (shares, commercial real estate, residential real estate etc.) can escape regulation of the central bank and institutions of prudential control for a long time, and can transmit effects on the rise in other prices, i.e. labour, utilities, etc. In such circumstances the credit channel of monetary transmission obtains special significance in the analysis of factors of destabilisation of the real economy. In accordance with the balance sheet channel of monetary transmission, higher and overvalued share prices enable greater borrowing by businesses at seemingly lower risks. Similarly, higher and overvalued real estate makes the balance sheet position of the company, the creditworthiness of citizens and other owners of assets look better. Banks underestimate risks and increase the supply of credit guaranteed by nominally more valuable or even overvalued collateral. The exposure of the financial system to risks increases unnoticeably. With rising concentration on collateralised or mortgage-based crediting the system grows more sensitive to asset price movements due to which smaller disruptions may cause major shocks and big disruptions may cause panic.

The fall in the liquidity of the system, as well as the fall in the liquidity of particular assets speeds up psychological effects which stimulate irrational behaviour of investors and speculators. In an attempt to obtain the free lunch, arbitrageurs additionally heat up the instability of the market through short sales, and the so-called noise traders do not allow possibly undervalued stocks or real estate to return to the fundamental realistic level.

We are witnesses of various forms of turbulences in the financial and real estate market as well as business cycle shifts in particular segments of national economies. High volatility of market interest rates, appreciation and then correction of share prices, appreciation and correction of real estate prices, structural causes of inflation, infection and spreading of infection by means of financial crisis, increasingly distinct process of price corrections of consumer goods caused by shocks in the market of energy products and agricultural products all represent major forms of instability which are characteristic of the development of the modern financial system. In different times and with different intensity, instability hits particular economies, and due to greater integration of global financial markets it is becoming a global problem.

Increasingly important non-fundamental factors as determinants of excess volatility of asset prices or misevaluation of assets are the reason for which asset price movements can become an independent factor of stimulating economic instability. In that case, the market is certainly ahead of the economy and dictates future possibilities of economic progress, stagnation or slowdown.

Literature does not provide a single answer to the following question: "Is the economy ahead of the market or is the market ahead of the economy?" In other words, "Do asset prices reflect the state of the economy or instead dictate its future?" At the same time, there are numerous examples which prove the accuracy of both the former and latter attitude through the existence of a two-way cause-effect relationship between: 1) trends in the economy and present and future asset price movements, and 2) asset price movements and the existing and future state of the economy.

The advocates of technical analysis have pointed out the existence of particular forms of behaviour of buyers and sellers of shares in the market depending on the historical price movement of stocks, changes in the overall trading volume, trading volume of particular shares, liquidity of a particular share and market liquidity, and impact of changes in the mood of buyers and sellers in the market. Graphs of share prices are full of rising and falling historic trends in which it is possible to perceive time points in which investor sentiment changes due to which, for example, during the upward trend sellers prevail over buyers and thereby change the past prevailingly upward trend into the new prevailingly falling trend.

Observed from the fundamental side, nothing important had to change, but investors under the influence of non-fundamental factors decided to change the direction of trends. It is interesting to notice a related fact that under those circumstances most investors sell particular shares because they think that in the future period their price will fall, and not because they think that those shares fundamentally are not worth as much as their current price is (if not why would they keep them in the first place).

4. PUBLIC EXPECTATIONS AND THE ROLE OF MONETARY POLICY

Monetary growth is the trigger of business cycles related to upward and downward trends in the economy. Functioning as the financial accelerator, the growth of money supply and the fall in (short-term and long-term) interest rates contribute to income growth and expansion of the economy. In the downward trend of the economy, the fall of the money supply arisen from panic in the financial sector stimulates additional effects of deepening recession (the opposite effect of the financial accelerator).

In a similar way as the liquidity effect is favourable to positive movements in the economy and increases in asset prices above their fundamental value, after the culmination of price heating, the return real income effect causes a series of unfavourable movements of the opposite trend. Namely, structural disruptions concealed over the years, unfavourable shifts in the business cycle of the economy and unsustainable growth based on monetary illusion cause reversible processes in time in terms of the fall in share and real estate prices, and lead to the disruption of stability of the financial market. Subsequently, the fall in real estate prices affects the development of an unfavourable business climate which deepens recession.

The fall in the value of asset prices weakens the capital position of the creditor (banks) and the balance sheet position of the borrowers (businesses). As a result, credit activities of banks are reduced, the money supply is contracted, investments and spending decrease, interest rates grow etc. In those circumstances the growth of interest rates is especially spurred by the incorporation of risk premium in the price of money since recession pressures increase the probability of moral hazard in terms of debtors' duty to honour credit obligations.

The fall in real estate and stock prices reduce household wealth which is reflected on decreased spending. Since personal consumption in the modern economy is the main generator of GDP changes, huge unfavourable multiplicative effects on the economy appear. Companies whose goods remain unsold and whose market value of stocks falls on a daily basis, face the problem of inability to provide fresh money in the banking and broader financial market. At the same time, banks face increased risks, fall in the credit quality (due to the decreased value of collateral and reduced debtor creditworthiness), by writing off bad loans etc. As a result, banks are getting more cautious when they grant credit--"credit is becoming more accessible to those who almost do not need them". Without the intervention of the central bank, this situation can produce powerful effects on the growth of long-term interest rates, plummeting share and real estate prices, falling investments and spending, decreased government tax revenues, growth of moral hazard, reduction and even negative net cash flow of businesses, development of debtor-creditor crisis or other unfavourable effects on the economy.

The inevitable effects of insecurity affecting public expectations are a very important factor through which the effect of expansionary or restrictive measures of monetary policy, as well as liquidity abundance or shortage, is reflected on business cycles. The uncertainty of future price movements of goods and services and asset prices changes business decisions of companies reduces interest in long-term investments and increases interest in short-term investments. Furthermore it affects the productivity of factors of production, conclusion of long-term agreements on future deals (delivery, construction of new housing facilities ...), tax decisions, savings preferences and household spending, as well as the need for risk protection instruments etc. The maintenance of price stability (prevention of high inflation and prevention of deflation) therefore represents a primary long-term goal of monetary policy. With rising importance of asset prices as the trigger of business cycles, another important long-term goal of monetary policy represents the stability of the financial system. The last is particularly emphasized in the research of Stiglitz and Greenwald (2003).

By stimulating the growth of interest rates during the period of the boom of stock and real estate prices, and by stimulating the fall in interest rates during the period of falling asset prices, monetary policy can act against the rise of panic in the financial markets. In addition, the policy of targeting inflation increases the credibility of monetary policy measures in the provision of stable macroeconomic frameworks. In accordance with that, Bernanke and Gertler (2001) conclude that due to reduced inflation expectations and due to the appearance of other psychological factors of inflation, the growth of asset prices causes at least minor inflation effects on the price of consumer goods in the circumstances of flexible targeting of inflation.

Literature provides conflicting attitudes on the responsibilities (of the agents) of monetary policy for the behaviour of asset prices, which implies conflicting attitudes on the benefit of actions of the central bank in preventing shocks in the share market and real estate market or the use of monetary policy for the purpose of stimulating shifts in the business cycle through the effect on asset price correction. Recent practice of the Fed to cut the official interest rate proves that the central bank is interested in preventing the prolongation of recession pressures and financial instability connected with the American mortgage market, as well as in preventing a further fall in prices as a result of which bear signals dominate in the share market.

Similar Fed interventions were noticed in earlier years related to the movements in the stock market and movements in the real estate market. Recession pressures in the USA during the 1990s were greatly caused by the fall in real estate prices, due to which the capital position of banks, as well as the balance sheet position of business borrowers was weakened. Similar cause-effect links between economic movements and asset price movements and future trends in the economy can be seen in the American economy since July last year, but with far more distinct and far more unfavourable effects on the economy.

5. ANALYSIS OF INTERDEPENDENCE OF MONETARY POLICY MEASURES AND ASSET PRICE MOVEMENTS--THE CASE OF CROATIA

When Croatia introduced the anti-inflation programme in 1993, it focussed its monetary policy on maintaining price stability using the anchor of nominal foreign exchange rates as the key instrument for achieving this goal. Interest rates on credits to businesses and households halved in the period between 1994 and 1999 (in 1994 they were about 25%), and continued the falling trend after that period. At the end of 2007 interest rates on short-term loans to businesses were 6.80% on average and to households 6.70%. This means that the liquidity of the system increased by relying on several sources: a) a more expansionary monetary policy of the central bank and credit policy of commercial banks, b) borrowing abroad by commercial banks and other sectors, c) income from the sales of public corporations to foreign buyers.

It is important to emphasise that in the period from 1996 until the end of 1999 Croatia suffered from a severe illiquidity problem, which was primarily the result of a restrictive monetary policy. The M1 monetary aggregate in the period almost froze at the level of HRK 13 billion. However, interest rates in that period were falling because of a great number of businesses which went bankrupt and because of reduced demand for money and low GDP growth rates.

At the beginning of 2000, monetary policy changed and became more expansionary, so that the M1 money supply surged by 40% and in the following two years by 30%. After 2000, primary money recorded significant growth due to the impact of the inflow of foreign currency related to the entry of foreign capital from ownership investments, foreign borrowing of the state, and later, foreign borrowing of banks etc. The growth of foreign currency assets in the balance sheet of the central bank is thus the key generator of the powerful growth of primary money and a key generator of the effect of expansionary measures of monetary policy. Although the M1 monetary aggregate growth rates indicated great expansion of the money supply, a big part of primary money was shrunk after--by the extremely high rates of the required reserve of 17% were imposed by the central bank. The final result was the increase in liquidity of the entire system and further fall in interest rates.

[FIGURE 1 OMITTED]

Apart from a more expansionary monetary policy, a significant impact on the increase in liquidity came from foreign direct investments which produced important monetary effects, i.e. they increased the liquidity of the system. From 2000 until the end of 2007, foreign direct investments in Croatia were about EUR 12bn.

The increase in liquidity and constant fall in interest rates since 1994 until today influenced the shifts in the real sector, in the money market, capital market and real estate market. The fall in interest rates encouraged the household sector to contract more housing loans which increased the demand for housing in the market thus causing a significant growth of the construction sector. See Figure 1. Growing income in the construction sector also increased the profit of business owners, which additionally triggered investments. Moreover, salaries rose in this sector which in turn increased demand in other sectors of the economy. Data and trends from the table 1 are even better visualised in the figure 2.

[FIGURE 2 OMITTED]

Since 1991 when Croatia became an independent state, the real estate market has been growing, and during the observed period not a single instance of fall in prices was noticed at any time. Demand for real estate is primarily the result of a more expansionary credit policy of banks, lower housing credit interest rates, the growth of household savings and constantly growing demand by foreigners for real estate, mainly on the Adriatic coast. The unspoiled healthy environment, the clean sea, excellent climate and expectations of rising prices on the Adriatic coast, but also in the continental region due to the anticipated entry of Croatia in the EU additionally heated the growth of real estate prices. As an example, 85% of Croatian citizens own a flat or a house, which next to Hungarian and Slovenian citizens, represents the highest percentage of ownership over this type of real estate in Europe.

It is interesting to notice that real estate investments in the capital of Zagreb were mainly the result of real housing needs, whereas real estate investments on the Adriatic coast (Jadran) were mainly speculative in character and related to the expectations of further growth of prices for all the mentioned reasons.

[FIGURE 3 OMITTED]

The analysis of trends in the Croatian real estate market can be done from the perspective of costs of undertaking the housing activity which typically involves the acquisition of a house or flat for residential purposes. These costs of undertaking the housing activity depend on the purchase price per [m.sup.2] of the housing real estate, on interest rates on housing and/or mortgage loans, on the expectations of growth of the real estate value and on the rate of depreciation of real estate value. In the past years interest rates on housing loans were significantly reduced which, together with lower costs, helped maintain the continuous growth of demand for housing real estate, especially newly built flats.

The figure 4 shows the process of increase in prices of newly constructed flats in the period from 2003 until the end of the first half of 2007. The first three columns show movements in prices of new flats under market conditions, and the separate fourth column shows movements in prices of new flats in the public system of subsidised housing projects (ATS). In relation to the first half of 2006, the prices of new flats outside the state programme of subsidised housing rose on average by 25.8% until the end of the first half of 2007. In the same one-year period, the prices of new flats in the capital Zagreb rose by 47%. The increase in real estate prices was somewhat affected by increased construction costs, and largely by rising costs of land where real estate was to be constructed and by utility costs. The rise in the prices of land aimed for construction in the most attractive parts of Croatia led to a significant divergence in price movements of real estate in different locations. Areas in which people do not want to live and build new houses and flats reflect stagnating or falling prices of land aimed for construction and housing facilities. At the same time, in big city areas (especially in the capital of Zagreb) where people want to live, available land for new construction is scarce and constant appreciation pressures appear in the market.

[FIGURE 4 OMITTED]

As a result of such scarcity and a very high price of construction land on attractive locations in big cities, construction investors move construction to less attractive suburban locations, which normally slows down the increase in prices of construction land in big cities. Interestingly, prices of flats in the observed period rocketed even in these less attractive city locations.

Table 2 shows percentage shares of different components in the forming of average price of 1[m.sup.2] of a new flat. Data for the capital Zagreb, other settlements and state subsidised housing are shown separately. Analysis shows shares of three separate categories:

1) construction land costs (purchase of land and possible costs of dislocation of tenants from the existing buildings if the construction site was not bought vacant),

2) profit of construction companies and construction costs (construction of a building includes tearing down existing buildings, cleaning the site, preparing the ground, elevating the building and roofing construction, installations and final works on the building/ flats), as well as

3) other costs (obtaining a construction permit, planning the building, measuring the land, professional supervision, different fees and taxes and insurance).

In contrast to this long period when real estate prices were going up in the Croatian market, the second half of 2007 shows the first signs of stagnation of real estate prices in the market and indicators of falling sales of new flats. The rise in real estate prices is still seen only on the Adriatic coast whereas Zagreb as a once attractive location records occasional periods of the fall in the real estate price index. See the figure 5. Unfavourable trends are especially observed in the price movements of houses in the Zagreb area.

At the beginning of 2006, and especially during 2007, the central bank made monetary policy measures stricter because of the rise in external debt of the country which seriously threatened to harm its foreign liquidity. In that sense, the central bank started to implement more fiercely a distinctly restrictive monetary and credit policy. Thus the credit activity of banks was linearly limited to a 12% rise a year, and any exceeding of this limit was punished by the purchase of compulsory Croatian National Bank Bills, or by less attractive immobilisation of bank money. Such non-selective attitude on decreasing bank credit activities might lead to the reduction in the amount of housing loans which would in return affect demand for flats and houses, the fall in prices and a downturn of the business cycle in the construction sector which significantly accounts for the GDP growth.

In addition, the fall in real estate prices might lead to adverse effects in the balance sheets of banks which granted housing loans taking under mortgage the same real estate whose prices, as was already mentioned, kept growing for the past 17 years. Any downward correction might produce lesser or bigger turbulences in banks.

[FIGURE 5 OMITTED]

An additional problem which the economy encountered at the beginning of this year is the significant rise in consumer prices which is primarily the result of external shocks: rise in oil and food prices in the global market. If the central banks keeps insisting on price stability as it announced, monetary and credit policies will have to be even more restrictive. This will lead to the rise in interest rates which will reduce investments and threaten with prorecession tendencies.

In that sense the central bank should treat housing loans differently and allow higher growth rates. On the contrary, more severe standstills in the construction sector might occur which means that this segment will contribute less to the GDP growth. Thus the asset channel involving real estate, if supported by a more flexible credit policy of banks, would yield better results in comparison to central bank regulation of price stability and liquidity of the system through interest rates.

The other transmission channel within the asset channel includes investments in securities, primarily shares in the capital market. In the period of transition, in which the Croatian economy and society are currently found, the capital market has been recognised primarily as a place where quick and easy profits can be made. Lack of experience and knowledge drove individual investors to make decisions on the basis of technical analysis of movements of future prices of securities. In that environment, stock market capitalization and stock index Crobex kept soaring. For example, stock market capitalization in 2006 doubled in relation to 2005. See the table 3. This definitely confirms the thesis that investment habits of individual but also institutional investors change, and that the bankocentric financial system started deconstructing more and more into a polycentric system. In addition, it is clear that the central bank started following shifts in the capital market with more intensity due to the growing influence of this segment of the financial market on the shifts in liquidity and demand for money. The asset structure changes increasingly by the year and gives room to more liquid forms such as securities.

The most important role in the increase in prices and transactions in the capital market was played by expectations heated up by different media. Two IPOs of two major Croatian companies "INA" (oil company) and HT (telecommunications) played a key role in promoting the shareholding culture. Encouraged by high yields from INA's IPO, investors (primarily individual) significantly increased interest in HT share investments. At the end of September 2007 citizens replied to HT's IPO by offering to purchase HRK 12bn (EUR 1.7bn). This means that citizens wanted to buy double the size of shares that the state offered as the co-owner of HT.

In a very short period (during September), HT's IPO caused disturbances in all segments of the financial market and monetary policy. Citizens exchanged their foreign currency savings for the domestic currency which created a pressure on the appreciation of the domestic currency. In the money market interest rates grew significantly due to withdrawals of money from banks. Banks felt a strong pressure on liquidity and looked for additional sources of funds. Moreover, the prices of other shares started to fall at that time. So, after the end of HT's IPO, the situation calmed and the rest of HRK 6bn mostly returned to banks. After this IPO share prices started to fall on the stock exchange so that for example the Zagrebacka bank Index fell by 14% during January 2008. Individual investors did not heed the warnings of leading economists or the governor of the central bank who all said that share prices were overvalued. Quite the opposite, during euphoria in the capital market during these two IPOs investors entered the stage of increased irrationality in terms of investment behaviour as they were dazzled with unrealistically high expectations about the yields.

So, restrictive monetary policy which was additionally made stricter throughout 2007, did not affect the fall in share prices at all, but quite the contrary, share prices recorded the highest increase. For the first time citizens took loans in banks to purchase shares (buying on margin) which in 2007 reached about HRK 600 million. That is why banks needed additional sources of funds therefore effectuating a rise in interest rates. This shows the rising importance of the financial market, shares in particular, in structuring assets of individual but also institutional investors and it also highlights their influence on the liquidity of the system and monetary policy. Therefore the asset channel of monetary transmission in Croatia has grown more important so that monetary policy and the central bank need to pay more attention to it than before. In conclusion, the real estate market and the capital market have become important segments in Croatia which affect the demand for money which further affects interest rates, monetary policy, liquidity and processes in the sphere of the real economy.

6. TESTING OF HYPOTHESES AND MODELS

In order to be able to observe the capital market or the movement of prices of real estate more closely multiple linear regression was used. Two models were made.

In model 1 the dependent variable is the price of newly built flats per [m.sup.2] on the level of average in the Republic of Croatia. Independent variables included the M1 money supply, interest rates, foreign assets, and the index of real effective exchange rates and allocated required reserve.

According to the model, if the M1 money supply increases by HRK 1 million then we can expect increases in real estate prices by 0.029 HRK, with other variables unchanged. The variable of money supply has no significant influence on real estate prices. The explanation for this can be found in the shifts of monetary multiplier at the level near 1, which reflects strong flows of shrinking of monetary effects of rise in primary money and simultaneous control of money supply growth. Furthermore, due to a high degree of Euroisation of the financial system and the economy, a large part of citizens' transactions relating to the purchase of real estate (especially from own deposits and without credits or borrowing) is effectuated in foreign money not included in the money supply.

If the interest rate increases by 1 percentage point then we can expect that the real estate price will rise by HRK 93.1649, with other variables unchanged. The interest rate variable is significant in the model (borderline). The result (contrary to the assumptions of the theory) can be explained with the influence of expectations of further growth of real estate prices because of which the rise in the price of borrowing increases demand for real estate, and its price. Furthermore, the volume of monetary transactions involving the purchase of real estate exceeds amounts of received housing mortgage credits, which implies that part of the purchase of real estate is financed from savings, and not from borrowing.

If foreign currency assets increase by HRK 1 million, then we can expect that real estate prices will increase by HRK 0.00479, with other variables unchanged. Foreign currency assets have no significant influence in the model.

If the index of real effective foreign exchange increases by 1 point then we can expect that the price of real estate will increase by HRK 195.37 with other variables unchanged. The variable of the real effective foreign exchange index is significant in that model. Real appreciation of Kuna over the years is linked with the continuous and growing inflow of foreign currency in the name of foreign debt, foreign investments, privatisation etc., which represents main flows of incentives to the liquidity effect and thereby increased demand for real estate.

If the allocated required reserve increases by HRK 1 million, then we expect real estate prices to grow by HRK 0.127, with other variables unchanged. The required reserve is significant in the model. The growth of allocated required bank reserve is linked with more banks borrowing abroad and increased nonresidential deposits and other sources of bank funds. Larger sources of funds imply (in terms of allocation of the required reserve) higher credit potential of banks.

The model is significant as a whole because the p-value (prob F-statistics) is smaller than 0.05. Variables INTERESTRATE, INDREALFX, RR are significant (at the level of 5% significance). The determination coefficient indicates that the model adjusts well to data. The Durbin-Watson ratio is at level 2 which shows that there is no autocorrelation of the first order. With the analysis of residuals all Q test values with divergence of 8 periods are non-significant (at the level of 5% significance).

In Model 2 the dependent variable is shifts in the Crobex share index. Independent variables include the money supply, interest rates, loans to households, foreign currency assets and the allocated required reserve.

According to the model, if the money supply increases by HRK 1bn then we can expect the reduction of Crobex by -0.0999 points, together with other variables unchanged. The money supply variable has a significant influence on prices (p-value <0.05).

If the interest rate increases by one percentage point then we can expect Crobex to reduce by 63.65 points with other variables unchanged. The variable of interest rate is significant in the model (border case). All of the mentioned reflects the influence of restrictive monetary policy measures on share prices.

If loans to households increase by HRK 1 m, then we can expect Crobex to grow by 0.0117 points, with other variables unchanged. The loans to households variable has a significant influence in the model.

If foreign currency assets increase by HRK 1 million then we can expect Crobex to rise by 0.008 points, with other variables unchanged. Foreign currency assets have no significant influence in the model.

If the allocated required reserve increases by HRK 1m, then we can expect Crobex to fall by 0.212 points, with other variables unchanged. The required reserve is significant in the model. All this reflects the influence of restrictive monetary policy measures in the share market.

The model is significant as a whole because the p-value (prob F-statistics) is smaller than 0.05. Variables INTERESTRATE, MONEYSUPPLY, HOUSINGLOANS and RR (required reserve) are significant with the level of significance of 5%. The determination coefficient indicates that the model adjusts well to data. The Durbin-Watson ratio is at level 2 which shows that there is no autocorrelation of the first order. With the analysis of residual all Q test values with divergence of 8 periods are non-significant at the level of 5% significance.

7. CONCLUSION

On the basis of the model and analysis of obtained results it can be concluded that the strongest influence on movements in the real estate market and capital market is effectuated by real effective foreign exchange rates, interest rates and required bank reserve. This is understandable because monetary policy aiming to reach and maintain price stability relies on foreign exchange rates and high rates of required banks reserves which neutralise the adverse effects of excess money in circulation.

On the other hand, it is interesting to notice that at the time of a highly restrictive monetary policy (although relative) during 2006 and 2007 the highest growth of share prices was recorded as well as the significant growth of real estate prices. Such shifts can be explained by a powerful effect of behaviourist factors on the movements of share prices and the appearance of euphoric optimism which spread in the public ignoring the warnings of agents of monetary policy who talked about the possibility of correction. By exhausting the liquidity effect over the years, correction and the real effect of money occurred.

According to the data of the Central Depositary Agency, at the end of 2007 Croatian citizens held about HRK 52 million in securities (predominantly shares), which accounts for 50% of savings deposits in banks. This reflects a change in the asset structure of households, and sends clear signals to the central bank on the major role that the asset channel plays in the monetary transmission mechanism.

REFERENCES:

Azali, M., "The Roles of Money and Credit in the Monetary Policy Transmission Mechanism: Preliminary Evidence Using Malaysian Quarterly Data", Asian Economic Review, Vol. 40(3), 1998, 395-405.

Bernanke, B., and Gertler M., "Should Central Banks Respond to Movements in Asset Prices?" American Economic Review, Vol. 91 (2), 2001, 253-57.

Dornbusch, R. and Fischer, S., Macroeconomics, McGraw-Hill, Inc., 1987.

Friedman, M., Schwartz, A. (1963), "Money and Business Cycles", Review of Economics and Statistics, No. 45, pp. 32-64

Fama, Eugene F., et al., "The Adjustment of Stock Prices to New Information," International Economic Review, Vol. 10 (1), 1969, 1-21

Gabe J. De Bondt, Financial Structure And Monetary Transmission In Europe: A Cross-country Study, Edward Elgar, 2000.

Ivanov, M. and Cavrak. V., "The Credit Channell of the Transmission Mechanism in the Republic of Croatia", International Journal of Entrepreneurship and Small Business, Vol. 2 (3), 2005, 254-265.

Mishkin Frederic S., "Housing and the Monetary Transmission Mechanism", Finance and Economics Discussion Series, Divisions of Research & Statistics and Monetary Affairs Federal Reserve Board, Washington, D.C., 2007, 1-55.

Mishkin, Frederic. S., The Economics of Money, Banking and Financial Markets, Pearson & Addison Wesley Publishing, New York, 2004.

Lovrinovic, I.; Benazic, M. "A VAR Analysis of Monetary Transmission Mechanism in the European Union", Zagreb International Review of Economics and Business, Vol. 7 (2), 2004, 27-42.

Lusardi, A., "Current Income, Permanent Income, and Consumption: Evidence from Two Panel Data Sets," Journal of Business and Economic Statistics, Vol. 14 (1), 1996, 81-90.

Souleles, N., "The Response of Household Consumption to Income Tax Refunds," American Economic Review, Vol. 89 (4), 1999, 947-58.

Stiglitz, E. J. and Greenwald, B., Towards a New Paradigm in Monetary Economics, Cambridge University Press, Cambridge, 2003.

Tobin J., "A general equilibrium approach to monetary theory", Journal of Money Credit and Banking, Vol 1 (1), 1969, 15-29.

Croatian National Bank, Monthly Report, December 2007, www.hnb.hr

Central Bureau of Statistics, Monthly Report, December 2007., www.dzs.hr

CentarNekretnina.net., www.centarnekretnina.net.

Zagreb Stock Exchange, www.zse.hr.

Dr. Marijana Ivanov earned her Ph.D. at the University of Zagreb in 2001. Currently she is a associate professor of Monetary Policy and Head of Department of Finance at Faculty of Economics and Business.

Dr. Ivan Lovrinovic earned his Ph.D. at the University of Zagreb in 1995. Currently he is a professor of Monetary Policy and International Financial Management at Faculty of Economics and Business. In the period 2003-2006 he was Dean of Faculty of Economics and Business at University of Zagreb.
TABLE 1: COMMERCIAL BANKS' LOANS TO HOUSEHOLDS

           Loans to to                            GDP at current
           households (in    Housing loans (in    prices
Period     million HRK)      million HRK)         (in million HRK)

1998.          17.672,50                               137.604,00
1999.          19.186,10             7.469,50          141.579,00
2000.          23.242,10             8.257,80          152.519,00
2001.          30.062,10             9.450,00          165.639,00
2002.          42.976,60            12.363,40          181.231,00
2003.          54.819,30            16.896,20          198.422,00
2004.          64.977,20            21.397,90          214.983,00
2005.          78.162,40            27.571,10          231.349,00
2006.          95.307,60            36.927,30          250.590,00

           Household         Housing
Period     loans to GDP      loans to GDP

1998.              12,84
1999.              13,55                 5,28
2000.              15,24                 5,41
2001.              18,15                 5,71
2002.              23,71                 6,82
2003.              27,63                 8,52
2004.              30,22                 9,95
2005.              33,79                11,92
2006.              38,03                14,74

Source: Data by Croatian National Bank, Monthly Report, December 2007.

TABLE 2: PERCENTAGE SHARES OF COMPONENTS IN FORMING OF AVERAGE PRICE OF
1M (2)

Location: The Capital--Zagreb

               Construction    Profit of construction companies   Other
                land costs         and construction costs         costs

   2003.          23,76                     53,68                 22,56
   2004.          19,14                     59,23                 21,63
   2005.          19,65                     64,99                 15,36
   2006.          15,73                     57,01                 27,26
I.-V1.2007.       15,36                     58,5                  26,14

Location: other settlements

               Construction    Profit of construction companies   Other
                land costs         and construction costs         costs

   2003.          16,37                     65,66                 17,97
   2004.          15,59                     65,03                 19,38
   2005.          14,99                     67,02                 17,99
   2006.          12,56                     65,9                  21,54
I.-V1.2007.        9,96                     68,33                 21,71

State subsidised housing

               Construction    Profit of construction companies   Other
                land costs         and construction costs         costs

   2003.           ...                      ...                   ...
   2004.           ...                      ...                   ...
   2005.           ...                      ...                   ...
   2006.           7,69                     72,47                 20,53
I.-V1.2007.        7,19                     73,17                 19,63

Source: Data by Central Bureau of Statistics, Monthly Report, December
2007.

TABLE 3: MARKET CAPITALIZATION ON ZAGREB STOCK EXCHANGE

                 Stocks             Bonds

1998.          18.275,80              -
1999.          19.225,10              -
2000.          22.178,60              -
2001.          25.815,50           7.199,80
2002.          28.320,40          10.131,00
2003.          37.130,50          13.418,80
2004.          61.734,40          24.563,70
2005.          80.724,50          34.399,40
2006.         161.692,20          40.011,30

Source: Data by Zagreb Stock Exchange, December 2007.

TABLE 4: TESTING OF HYPOTHESES--MODEL 1

Variable        Coefficient    Std. Error     t-Statistic     Prob.

C              -13747.73       5249.772      -2.618729        0.0589
MONEYSUPPLY     0.029150       0.035866       0.812764        0.4620
INTERESTRATE    93.16498       40.58385       2.295617        0.0833
FCASSET         0.004790       0.017788       0.269270        0.8010
INDREALFX       195.3716       43.90222       4.450153        0.0112
RR              0.127132       0.033830       3.758008        0.0198

R-squared 0.923762

Adjusted R-squared 0.912368

S.E. of regression 181.658

Sum squared resid 131998.6

Log likelihood -61.62919

Durbin-Watson stat 2.032421

Mean dependent var 8455.400

S.D. dependent var 613.6552

Akaike info criterion 13.52584

Schwarz criterion 13.70739

F-statistic 19.74054

Prob(F-statistic) 0.006380

Source of data: Croatian National Bank, Central Bureau of Statistics,

TABLE 5: TESTING OF HYPOTHESES--MODEL 2

Variable        Coefficient    Std. Error    t-Statistic    Prob.

C               2281.434       575.7356      3.962642       0.0166
MONEYSUPPLY    -0.099954       0.034187     -2.923703       0.0431
INTERESTRATE   -63.65250       23.17253     -2.746895       0.0515
HOUSINGLOANS    0.117386       0.020935      5.607114       0.0050
FCASSET         0.008141       0.011551      0.704758       0.5198
RR             -0.212283       0.042279     -5.020978       0.0074

R-squared 0.857564

Adjusted R-squared 0.957036

S.E. of regression 126.9369

Sum squared resid 64451.90

Log likelihood -58.04483

Durbin-Watson stat 1.993452

Mean dependent var 1248.304

S.D. dependent var 612.4000

Akaike info criterion 12.80897

Schvwarz criterion 12.99052

F-statistic 41.09548

Prob(F-statistic) 0.001565

Source of data: Croatian National bank, Central bureau of Statistics,
Zagreb Stock Exchange
Gale Copyright:
Copyright 2008 Gale, Cengage Learning. All rights reserved.