Understanding the Role of Financial Markets in Business Cycles: Financial Accelerator, Study notes of Economics

Download Understanding the Role of Financial Markets in Business Cycles: Financial Accelerator and more Study notes Economics in PDF only on Docsity! B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 171The financial accelerator effect: concept and challenges BRUNO ĆORIĆ, PhD*1 University of Split, Faculty of Economics, Split bcoric@efst.hr Review article**2 JEL: E32, E44 UDC: 336 * This paper is based on the author’s PhD dissertation at Staffordshire University. I thank Nick Adnett, Geoff Pugh, Peter Reynolds, Bill Russell, Ahmad Seyf and two anonymous referees for their many helpful sugge- stions. Any remaining errors are the responsibility of the author. ** Received: November 4, 2010 Accepted: April 28, 2011 B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 172 Abstract This review concentrates on the role of information asymmetry in fi nancial markets in the amplifi cation and propagation of short-run output fl uctuations. We fi nd that the fi nancial accelerator effect, as it is known, provides a consistent, fi rst princi- ple based, theoretical framework for the analysis of the relationship between fi - nancial markets and short-run output fl uctuations. It also provides a plausible explanation of the proximate causes of the recent crisis, and fi rst principle-based theoretical background for the credit policy measures taken during this crisis by many central banks and fi scal authorities. Despite the theoretical plausibility, the empirical evidence about the economic importance of the fi nancial accelerator effect is still relatively weak. We also suggest two new aspects to expand existing concept of the fi nancial accelerator effect, which call for further research. Keywords: asymmetric information, fi nancial markets imperfections, fi nancial ac- celerator, business cycles 1 int roduct ion In 1988 Mark Gertler summarized the stance of macroeconomic literature in the following sentence; Most of macroeconomic theory presumes that the fi nancial system functions smoothly – and smoothly enough to justify abstracting from fi - nancial considerations (Gertler, 1988:559). Twenty years after, the severe conse- quences of the worst fi nancial crisis since the Great Depression have drawn atten- tion of a broader economic community to the idea that episodes of deteriorating credit market conditions, growing debt burdens and falling asset prices in fi nan- cial markets are not just passive refl ections of a declining economy, but can them- selves be a major factor depressing real economic activity. The potential for the active role of fi nancial markets in real economic activity was recognized and ex- tensively investigated in the macroeconomic literature over the last two decades. Most of these researches have focused on the fi nancial accelerator effect. This literature essentially incorporates the partial equilibrium analyses about in- formation-based imperfections on fi nancial markets into a general equilibrium framework. The idea that the asymmetric information-based imperfections in fi - nancial markets can infl uence short-run aggregate economic activity was sugge- sted by Bernanke (1983). In particular, Bernanke (1983) argued that the credit squeeze during the great depression arose mainly from the autonomous reaction of the banking sector to increased real costs of intermediation caused by worse- ning asymmetric information problems in the fi nancial markets during this period. In other words, the credit squeeze was not just the passive response of decreasing demand for loans due to the decline in economic activity. Despite this early work, it was only after Bernanke and Gertler (1989) formalized this idea in a general equilibrium framework that it attracted the broader attention of the economic au- dience. Bernanke and Gertler (1989) constructed a general equilibrium model with incomplete fi nancial markets, where the actions of all economic agents pro- B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 175Second, in cases when the borrowers have superior information about projects’ characteristics (value, riskiness, etc.), or abilities to take unobserved actions that can affect projects’ return distribution, a greater incompatibility of interests between borrowers and lenders increases agency costs.5 What types of costs age- ncy costs would include depends on the way the information asymmetry in the credit market is modelled (see below). In brief, if a model is built on the assump- tion that information problems are solvable, then the agency costs will include costs of all real resources that the lender would be “forced” to spend to overcome these problems. On the other hand, if the underlying assumption is that those pro- blems are not solvable, or at least not completely solvable, and credits are rationed in equilibrium, then agency costs will include the costs of suboptimal allocation of funds in the economy. So, the greater the incompatibility of interests between lenders and borrowers, the greater the extent of real resources spent on monito- ring, selecting, etc., and/or the greater the extent of credit rationing in the eco- nomy. Consequently, the greater the incompatibility of interests, the higher the interest rate and/or the lower availability of credits. The fi nancial accelerator effect can be described as follows (fi gure 1). A change in aggregate economic activity causes a change in economic agents’ net worth because of a positive correlation between them. Due to imperfect information, the terms under which economic agents are able to raise external fi nance, hence also the external fi nance premium, are inversely related to their net worth. In so far as the external fi nance premium is inversely related to economic agents’ net worth, the procyclical behaviour of economic agents’ net worth over business cycles implies countercyclical behaviour of the external fi nance premium. This inverse relation between output changes and the external fi nance premium makes bor- rowing more diffi cult and/or expensive during recession than during the expansio- nary phase. This in turn exaggerates swings in investment, spending and produc- tion over business cycles. For example, any negative economic shock that might lead to a decrease in economic agents’ net worth would also increase the external fi nance premium. Consequently, due to higher costs, and/or reduced ability to borrow, the overall level of agents’ investments, spending and production will decrease. In turn, that will depress the economy even further. Economic disturbances that can be amplifi ed and propagated by the fi nancial ac- celerator mechanism include all shocks that cause: changes in the value of econo- mic agents’ liquid assets (change in cash position, short-term fi nancial asset, etc.), such as, a change in productivity, a change in aggregate demand caused by a de- crease in money supply, a decrease in foreign demand, etc.; changes in the value 5 Agency costs are a broad range of different costs that arise in all economic transactions where the parties involved in those relations have different interests and where their actions are not observable, or at least not costlessly observable by all transaction participants. They are the wedge between the so-called first best and the second best solution. That is, the difference between costs of the transaction that takes place between two parties in situations in which the information is equally shared and costless, and the costs of the same trans- action in the situation where information is not equally distributed between parties. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 176 of economic agents’ illiquid assets, that can arise due to either a change in interest rate or a change in the asset prices both of which can be caused by actions of eco- nomic policy makers or by changes in the expectations about future economic performance, etc.; change in economic agents’ outstanding obligations can also be caused by a change in the interest rate, under the assumption that outstanding loans are subject to variable interest rates. FIGURE 1 The fi nancial accelerator effect The arrows symbolize the mechanisms (channels) through which the financial accelerator opera- tes, and the boxes represent events and their consequences for the economy. Arrow 1 represents a positive relationship between changes in aggregate economic activity and agents’ net worth. In turn, Arrow 2 represents an inverse relationship between net worth and the size of the exter- nal finance premium. Arrow 3 represents an inverse relationship between the external financial premium and investment, spending and production. Finally, the return arrows represent pro-cyc- lical feedback into aggregate economic activity. 3 MODELLING STRATEGIES The different modelling approaches to the fi nancial accelerator effect refl ect: the variety of different situations in which the asymmetric information problem between borrowers and lenders may emerge; different fi nancial markets in which this problem can appear; different attitudes toward the question whether partici- pants in fi nancial markets can overcome the problem of asymmetrically distribu- ted information or not; and different types of borrowers whose economic activities can be infl uenced by fi nancial market imperfections. 3.1 DIFFERENCES IN MODELLING INFORMATIONAL ASYMMETRY In general, it is possible to distinguish three major types of modelling strategies used by researchers to incorporate the fi nancial accelerator effect into general equilibrium models (the main differences among these approaches are summari- sed at the end of this section in the table 1). The fi rst two approaches, established by Bernanke and Gertler (1989) and Kiyotaki and Moor (1997), consider informa- tional asymmetry on credit markets as the cause of the fi nancial accelerator effect. The third approach, established by Greenwald and Stiglitz (1993) considers infor- mational asymmetry on equity markets and managers’ risk aversion as the cause of this effect. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 177Differences between the fi rst two approaches, which concentrate on credit markets, arise due to the variety of situations in which informational asymmetry between ultimate savers and investors on the credit markets may be evident, and differen- ces in the underlying assumption taken by these authors with respect to the que- stion whether participants in fi nancial markets can overcome the problem of asymmetrically distributed information or not. In particular, the asymmetric information problem can emerge because borrowers’ intentions, or honesty, or project riskiness, or project quality, or any combination of these characteristics is not observable to lenders before transactions take place (ex-ante asymmetric information problems). Once the transaction has taken place, lenders may not be able to observe borrowers’ actions, project return or duration, or to force borrowers to repay loans (ex-post asymmetric information problems) (Jaffe and Stiglitz, 1990). A model that would include all the possible scenarios in which an asymmetric information problem may arise as sources of the fi nancial accelerator effect would be extremely complicated, very diffi cult to build and hi- ghly intractable. The strategy adopted by researchers has been to use one particu- lar formulation of asymmetric information problems as representative of all sce- narios and to assume a priori whether or not those problems are solvable. As a result, different models have different formulations of the agency costs and the external fi nance premium. The assumption that asymmetric information problems between borrowers and lenders are solvable implies that the lenders would use various techniques to deal with this problem: for example, screening; selecting among borrowers; monito- ring (inspection of borrowers’ cash fl ows, balance sheet position, management, realized return, etc.); engaging in a long-term relationship with borrowers; and enforcement of restrictive covenants such as a minimum solvency ratio or a mini- mum cash balance, etc. The use of all these instruments requires lenders’ real re- sources and entails real costs for which lenders have to be compensated in equili- brium. Therefore, even in the cases when these techniques are able to reveal all the hidden or unobservable borrower characteristics, their implementation makes ex- ternal sources of fi nance expensive to the borrower compared to the use of internal fi nance, where these costs are not present. The difference between these two sour- ces of funds will be expressed in the form of the higher interest rate borrowers will be asked to pay. The external fi nance premium will in this case take the form of the difference between the interest rate on the external source of fi nance that age nts with low net worth would be charged compared to the interest rate those agents would be charged in the situation where their net worth is high. On the other hand, the assumption that asymmetric information problems between borrowers and lenders are not solvable implies that the borrowers would not always be able to obtain credit, or at least not the amount considered optimal from their point of view. In particular, in circumstances when lenders are not able to overco- B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 180 when a project’s NPV is positive, fi rms will often be reluctant to issue new shares and invest, because the decrease in share prices can outweigh increase in value arising from the project’s positive NPV. Further, Greenwald and Stiglitz (1993) also assume that fi rms behave in such a manner as to minimize the probability of bankruptcy, since bankruptcy is costly for stockholders and especially for mana- gers, due to loss of reputation. These costs induce fi rms to act in a risk-averse manner. Finally, fi rms operate in a stochastic environment where the outcome of every production and investment decision is uncertain. In an environment which is intrinsically stochastic and in which fi rms are risk-averse, changes in fi rms’ net worth position can have potentially large effects on their willingness to produce. For example, economic shocks that reduce fi rms’ net worth also reduce the level of their own funds with which the fi rm may contribute to investment in production. Consequently, if fi rms want to maintain the same level of investment in produc- tion, they have to increase borrowing. Insofar as debt obligations are in the form of state-independent claims, a rise in the share of fi xed obligations in fi nancial sour- ces, in a stochastic world, increases the probability of bankruptcy. Since fi rms are risk averse they will be reluctant to increase borrowing, but will rather decrease investment in production. Decreases in fi rms’ production get translated into de- creases in the demand facing other fi rms, and through this mechanism shocks get transmitted further. Overall, similar to the previous models, fi nancial market im- perfections will, due to net worth procyclicality, amplify changes in economic ac- tivity over the business cycle. So far, this approach to modelling the fi nancial acce- lerator effect has been adopted by Arnold (2002) and Gatti et al. (2007). TABLE 1 Different approaches to modelling the fi nancial accelerator effect Bernanke and Gertler’s (1989) Kiyotaki and Moor’s (1997) Greenwald and Stiglitz’s (1993) Cause of the fi nancial accelerator effect informational asymmetry on credit markets informational asymmetry on credit markets informational asymmetry on equity markets and risk aversion of fi rms’ managers Ability of fi nancial markets participants to overcome informa- tional asymmetry problem can be solved problem cannot be solved problem cannot be solved Change in net worth induce change in cost of credits availability of credits willingness to borrow B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 1813.2 DIFFERENT TYPES OF BORROWERS Another feature of the fi nancial accelerator literature is that almost all studies in this fi eld concentrate on one group of economic agents.6 Following Bernanke and Gertler (1989), Greenwald and Stiglitz (1993) and Kiyotaki and Moore’s (1997) seminal contributions, the large majority of general equilibrium models in this fi eld concentrate on fi rms’ borrowing in the formulation of interactions between imperfect fi nancial markets and short-run economic fl uctuations. The general orientation of the literature toward explanation of the fi nancial accelerator effect through the behaviour of one group of economic agents (usually fi rms) is based on the intention to reduce technical concerns and keep models relatively simple to maintain tractability. This strategy is not a product of beliefs that the activity of other economic agents is not infl uenced by the same kind of fi nancial market im- perfections. On the contrary, as Bernanke et al. (1999) emphasized, a complete description of the fi nancial accelerator mechanism will likely include a signifi cant role for non-fi rm borrowers such as households and banks. The fi nancial accelerator effect on household spending occurs because house- holds, as well as fi rms, fi nance some of their expenditures by borrowing. In parti- cular, households usually fi nance investments in housing and purchases of other durable goods by raising funds in credit markets. These fi nance transactions are also characterized by asymmetric information problems between the borrowers (households) and the lenders (banks). Therefore, households’ ability and/or terms under which they are able to obtain funds, hence their spending, are also infl uen- ced by their net worth. Since empirically a large proportion of households’ bor- rowings are secured by real estate, the literature has focused primarily on the ef- fect of changes in house values (seminal contributions include: Aoki et al., 2002, 2004; and Iacoviello, 2005). Aoki et al. (2004) described the fi nancial accelerator effect on household spending as follows. A positive shock to economic activity causes a rise in house prices, which leads to an increase in homeowners’ net worth. This decreases the external fi nance premium, which leads to a rise in housing investments and also spills over into consumption demand (spending on dura- bles). Consequently, the same mechanism that propagates economic shocks through fi rms’ investments should also work through households’ spending and invest- ments decisions. It does not seem unreasonable to assume that the effect of the net worth changes on households’ investments should be even more pronounced than on the fi rms’, since households are unable to raise alternative sources of fi nance through equity or bond issues, but are exclusively oriented towards banks as the sources of external funds. Recently Gertler and Kiyotaki (2010) developed a general equilibrium model in which the fi nancial accelerator effect emerges due to an asymmetric information 6 For a model which simultaneously combines the financial accelerator effect on firms’ and banks’ economic activity in a partial equilibrium environment see Holmstrom and Tirole (1997). Iacoviello (2005) combines the financial accelerator effect on firms’ and households’ activity in a general equilibrium model. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 182 problem that constrains the ability of banks to obtain funds from depositors in retail as well as in wholesale (“inter-bank”) fi nancial markets. Since banks enter the deposit market as borrowers and given that they can go bankrupt as well as fi rms, there is no reason to assume that banks’ ability to collect funds and/or the costs of the funds will not be infl uenced by their net worth (bank capital). To the extent that the economic shocks affect banks’ net worth7 it might also affect banks’ ability to attract funds. Why does banks’ net worth matter for attracting funds? The reason is the same as before – information asymmetry. Depositors cannot observe how much risk banks are taking while investing their money. However, they know that bankruptcy is costly. Higher bank net worth implies larger costs of bankruptcy and, hence, lower incentive to take risk. Additionally, higher bank net worth means a more liquid bank, that is, a better ability to compensate unanticipa- ted losses and again a lower probability of bankruptcy. In sum, depositors will be more willing and/or will request a lower deposit interest rate to put their money in bank with high net worth. They will also be more willing to “contribute” to bank capitalization by investing in bank’s issue of new shares if that bank has higher net worth. This is important because a negative effect of a reduction in banks’ net worth on loans supply can also be caused by banks’ intention to maintain a desired or imposed capital adequacy ratio (see below). Unlike that of fi rms and house- holds, the economic activity of the banks does not consist of spending on durables and houses, or of production and investment. Yet, since banks fi nance these acti- vities to a large extent, a change in their net worth can have an effect on aggre gate spending, investment and production. The terms and amounts of the funds the banks collect directly determine the terms under which they lend, and the amo unts they are able to offer to potential borrowers. Therefore, adverse changes in their net worth can cause credit tightening and negatively affect households’ and fi rms’ expenditure and overall economic activity. 4 MONETARY TRANSMISSION MECHANISM The fi nancial accelerator effect has recently been used in many different areas of research, for example: Cepedes et al. (2004), Gertler et al. (2007) and Magud (2010), among others, analysed properties of different exchange rate regimes using the fi nancial accelerator framework; Olivero (2010) explored the relation- ship between the fi nancial accelerator and international transmission of business cycles; Aghion et al. (2005) incorporated this effect in the analysis of the relation- ship between output volatility and growth. Assuming that entrepreneurs raise funds on imperfect credit markets, Wasmer and Weil (2004) developed a theory of job creation and job destruction that suggests the existence of a fi nancial accelerator effect based on the general-equilibrium feedback between credit and labour market; Portes (2007), Portes and Ozenbas (2009), and Ćorić and Pugh (2011) explore the role of the fi nancial accelerator effect in the so-called great modera- 7 For example, negative economic shock can increase the non-performing loans rate and/or decrease the value of banks’ securities. In the case of the exogenously induced increase in interest rate banks can also suffer capi- tal erosion due to a maturity mismatch between banks’ assets and liabilities. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 185The fi nancial accelerator literature also offers a plausible fi rst principle based ex- planation of the proximate causes of the recent crisis. If we consider the collapse of the subprime mortgages market in September of 2007 as the initial economic shock, then this shock can hardly be categorised as large. Although the size of subprime mortgages market was large in absolute terms ($0.7 trillion) it was less than 0.5 percent of the size of US fi nancial markets (Bank of England, 2007). Hence, it is puzzling how default on these loans could become the source of the most severe crisis of this era. A possible explanation can be summarised as fol- lows. The collapse of the US subprime mortgage market forced fi nancial interme- diaries to write off hundreds billion dollars in bad loans and caused an erosion of their capitalization. Deterioration of intermediaries’ balance sheets harmed their ability to raise funds, which reduced their lending capacity. Consequently, lending standards that non-fi nancial borrowers face tightened and loan supply declined. The credit crunch adversely affected investment, consumption and property prices that are sensitive to the free fl ow of credits and lending standards. Decline in house and real estate prices caused erosion of households’ and fi rms’ net worth and ac- cordingly a decline in their debt capacity. Weakening of non-fi nancial borrowers’ debt capacity further increased external fi nancial premium and amplifi ed existing decline of investment, consumption and output. Decline of the aggregate econo- mic activity, the ensuing rise in unemployment and reduction in housing prices increased the amount of nonperforming loans. This reduced intermediaries’ profi - tability and further deteriorated their net worth with another pro-cyclical feedback into the aggregate economic activity. Overall, the initial shock to the fi nancial sector itself seems to have caused and/or intensifi ed the deterioration of the net worth of all three groups of economic agents and through the fi nancial accelerator effect generated a severe crisis.11 The fi nancial accelerator can also provide a theoretical background for the credit policy measures taken during this crisis by the Fed and many other central banks and fi scal authorities, which are often seen as old-fashioned and contradictory of modern economic theory. In particular, the Fed reacted to the current crisis using three types of credit policy measures: by making imperfectly secured loans to fi - nancial institutions; by lending directly to high grade non-fi nancial borrowers; and by equity injection and debt guarantees to large fi nancial institutions (Gertler and Kiyotaki, 2010). The same or similar measures to facilitate credit fl ows and boost fi nancial institutions’ net worth were taken by monetary and fi scal authori- ties of many other countries as well. Using the dynamic general equilibrium mo- 11 It is important to stress that this is just one of the possible explanations of proximate causes. The ultimate causes of the recent financial crisis are still a matter of debate and intensive research. At the moment a few questions seem to be crucial with respect to the ultimate causes of recent financial crisis. What was the source of excessive liquidity on the financial markets of the US and other industrialized countries in the years before outbreak of recent financial crisis? To what extent did financial innovations and financial derivatives, which had been supposed to make the financial system more stable by diversifying risk and rising liquidity, amp- lify the problem of information asymmetry? Why did financial institutions engage in excessive risk taking? To what extent might the principal-agent problem between the owners and managers of financial institutions contribute to excessive risk taking? B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 186 del in which the fi nancial accelerator effect emerges due to changes in banks’ net worth Gertler and Kiyotaki (2010) demonstrated that implementation of this kind of measures can have benefi cial effect on the aggregate economic activity during the crisis by reducing the rise in the external fi nance premium. The results of Ger- tler and Kiyotaki’s (2010) model simulations also demonstrated that the net bene- fi ts from these credit market measures increase with the severity of the crises, which helps to account for why it makes sense to employ them only in the cases of severe recession. These results are important because they demonstrate that these measures are in line with modern, fi rst-principle based macroeconomic theory. Theoretical consistency and the ability to offer a plausible explanation of the recent crisis has drawn attention of a broader economic community toward the idea of the fi nancial accelerator. However, despite its recent popularity in the broader econo- mic community, the empirical evidence on the economic signifi cance of this effect is still relatively weak. The empirical estimates of the fi nancial accelerator effect face three main diffi culties. Problems that are commonly acknowledged are: po- tential endogeneity that hampers any evaluation of aggregate data, and inability directly to observe and measure the external fi nance premium. A problem that has not been recognised, and that makes an empirical assessment of the fi nancial accelerator effect even more diffi cult is that the existence of a procyclical external fi nance premium is a suffi cient but not a necessary condition for the existence of the fi nancial accelerator effect. The fi nancial accelerator effect suggests that adverse economic disturbances lead to a decrease in aggregate investment due to a shift in the credits supply curve caused by increases in asymmetric information costs. Hence, it seems that empiri- cal tests can focus on the possible existence of procyclical movements in aggre gate credit fl ows, that is, the existence of a positive correlation between aggregate credit and aggregate output movements. The strong positive correlation between these variables is, in general, observed in aggregate time series (see for example: Del- l’Ariccia and Garibaldi, 2005). However, the joint test based on aggregate data is unlikely to be informative about the existence of fi nancial accelerator (Bernanke et al., 1996). This testing procedure suffers from the so-called identifi cation equiva- lence problem. The positive correlation between aggregate credit changes and GDP changes can be caused not just by movements in credit supply but by move- ments in credit (investment) demand as well. For example, King and Plosser (1984) demonstrated that procyclical movements of aggregate credit fl ows can be genera- ted by a frictionless real business cycle model, in which changes in aggregate cre- dits fl ows are caused solely by shifts in investment demand. The inability to distin- guish between different sources of observed colinearity makes this testing proce- dure unsuitable. Hence, the empirical literature has followed alternative identifi ca- tion strategies. These empirical investigations consist of two groups of studies. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 187The fi rst group of studies focuses on the relationship between variables that are meant to proxy the external fi nance premium and aggregate economic activity. That is, in the majority of models, the fi nancial accelerator effect is based on a countercyclical external fi nance premium that is not directly observable. Gertler and Lown (1999) argued that the development of a high-yield debt market (the “junk bonds” market) in the early 1980s in the US can potentially resolve this problem. According to Gertler and Lown (1999) a high-yield bond spread12 should provide a good overall indicator of the external fi nance premium since fi rms that raise funds by issuing high-yield bonds are the kind of fi rms that can be supposed to face frictions on credit markets. On the other hand, AAA rated fi rms are very unlikely to face asymmetric information problems on fi nancial markets. During recession the cost of fi nance (bond rate), according to the fi nancial accelerator effect, should increase more for fi rms that are more exposed to asymmetric infor- mation. Hence, the high-yield bond spread should rise. The reverse should be true over the economy expansionary phase. The results of their empirical tests confi r- med this prediction of the fi nancial accelerator theory. In particular, they found a strong inverse relationship between high-yield spread and output gap. The results of Gertler and Lown (1999) were confi rmed by Mody and Taylor (2004), who recorded a high negative correlation between high-yield spread and annual per- centage changes in real GDP. Recently, Aliaga-Diaz and Olivero (2010) found that price-cost margins for US banks are also consistently countercyclical, which can be an indicator of the fi nancial accelerator effect in the banking sector. Al- though these results are consistent with the theory of the fi nancial accelerator, a few issues challenge robustness and raise some concerns about this testing ap- proach. First, whether high-yield bond spread is good proxy for external fi nance premium is a matter of debate; AAA rated fi rms are on average considerably lar- ger than fi rms that raise funds on the high-yield bond market. The recent econo- mic crisis has demonstrated that size (still) matters. Hence, the evidence that the bond rates of AAA rated fi rms increase less during recession than bond rates on the high-yield bond market can refl ect investors belief in the “too big to fail” hypothesis. Furthermore, the large companies are usually more internationally di- versifi ed, which can make them less exposed to change in domestic GDP. There- fore, the negative correlation between high-yield spread and annual percentage changes in real GDP can also be the results of the larger international diversifi ca- tion of AAA rated fi rms. Second, identifi cation of a countercyclical external fi nance premium does not necessary imply that this premium intensifi es the effects of aggregate shocks. Third, the results of these studies are based only on US data, using a relatively short series, with only one signifi cant recession in the cases of Gertler and Lown (1999) and Mody and Taylor (2004). Fourth, due to the short time period and lack of data for other countries, the potentials for further applica- tion of this approach are limited. Finally, the existence of external fi nance pre- mium is not a necessary condition for the fi nancial accelerator effect. In particular, 12 The high-yield bond spread is the difference between high-yield bond rate and the corresponding rate for the bonds of AAA rated firms. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 190 reducing real value of borrowers’ outstanding debt obligations. In that respect adverse supply shocks can be benefi cial for the borrowers’ net worth and conse- quently “decelerate” an economic shock. With respect to possible further theoretical research, we suggest two new aspects to expand the existing concept of the fi nancial accelerator effect. Existing literature accounts for the problem of information asymmetry between economic agents, but not for the information asymmetry “within economic age- nts”. In particular, in many fi rms, especially in banks, the owners are not the ma- nagers. Asymmetric information between owners and managers results in what is called the principal-agent problem. This problem is potentially important for the fi nancial accelerator effect and its introduction in the fi nancial accelerator fra- mework can be informative. Namely, an economic agent’s net worth is important for lenders because high net worth implies that an economic agent will behave more diligently and/or take less risk since its losses in the case of bankruptcy will be larger. However, in a case in which managers are not the owners of a fi rm or bank, their losses in the case of bankruptcy (loss of reputation) seem to be inde- pendent of the fi rm’s or bank’s net worth. Therefore when the principal-agent problem exists, an economic agent’s net worth can lose its relevance for the len- der. It is possible to speculate that the fi nancial accelerator effect would still exist in this case since a higher economic agent net worth also implies a lower probabi- lity of bankruptcy, and also to speculate that in this case it is relative rather than absolute net worth that would be important to a lender. However, this should be investigated in more detail. The literature also presumes that only an individual economic agent’s net worth is important. The underlying premise is that a lender is able to observe a borrower’s net worth. This assumption can be easily justifi ed when banks are lenders. Howe- ver, it seems less justifi ed when banks are borrowers. Although banks are forced to release information on their capitalization on a regular basis it still remains doubtful how much fi rms, and especially households are informed about their net worth. For example, even if households do have information about these data re- leases it is very doubtful whether they have enough knowledge to understand them and compare capitalization among different banks. Consequently, it would be useful to analyse whether in the case of banks, information about soundness of the overall banking sector that can be communicated through the media or central banks’ announcements; what the function of banks’ aggregate net worth can be is also important. With respect to the results of existing fi nancial accelerator models this would not make difference since their mathematical formulation is based on the representative agent approach. However, this might be important for the ex- planation and especially for the empirical assessment of the fi nancial accelerator effect. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 1916 CONCLUSION The severity of the recent economic crisis raises a question about the role of fi nan- cial markets in modern market economies. This review concentrates on the rela- tionship between information asymmetry on fi nancial markets and short-run aggregate economic fl uctuations, the so-called fi nancial accelerator effect. We found that the fi nancial accelerator effect offers a consistent, fi rst-principle based, explanation of the relationship between fi nancial markets and short-run aggregate economic fl uctuations based on informational asymmetry on fi nancial markets. This effect also offers a plausible rationalization of the severe conse- quences of the subprime mortgages market’s crash in September of 2007. Finally, this effect, or more precisely, the prevention of its even stronger manifestation, provides a theoretical background for the credit policy measures taken during the recent crisis by many central banks and fi scal authorities. These features made the fi nancial accelerator effect recently very popular in the broader economic community. Despite its popularity, and the recent fi nancial cri- sis, which seems to confi rm its signifi cance, the empirical literature has faced se- rious challenges in the empirical identifi cation of this effect. Overall, we found that existing empirical literature is still unable to provide robust assessments of the size and economic relevance of the fi nancial accelerator effect. More empiri- cal research is necessary especially because the recent literature demonstrates that this effect theoretically could go in the opposite direction and reduce output vola- tility as well. Finally, we made a case for two new lines of enquiry. In particular, we found that it might be informative to introduce the principal-agent problem between owners and managers into the fi nancial accelerator framework; and also to consider the role of the aggregate net worth of banking sector in the fi nancial accelerator effect and its empirical estimation. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 192 LITERATURE AGHION, P. [ET AL.], 2005. “Volatility and Growth: Credit Constraints and Produc- tivity-Enhancing Investment”. NBER Working Paper, No. 11349. Aikman, D. and Vileghe, G., 2004. “How Much does Bank Capital Matter?”. Bank of England Quarterly Bulletin, 44 (1), 48-58. Aliaga-Diaz, R. and Olivero, P. M., 2010. “Is there a fi nancial accelerator in US banking? Evidence from the cyclicality of banks’ price-cost margins”. Econo- mics Letters, 108 (2), 167-171. Almeida, H. and Campello, M., 2007. “Financial Constraints, Asset Tangibility, and Corporate Investment”. The Review of Financial Studies, 20 (5), 1429- 1460. Almeida, H., Campello, M. and Liu, C., 2006. “The fi nancial accelerator: Eviden- ce from International Housing Markets”. Review of Finance, 10 (3), 321-352. Aoki, K., Proudman, J. and Vlieghe, G., 2002. “Houses as Collateral: has the Link between House Prices and Consumption in the U.K. Changed?”. The Federal Reserve Bank of New York Economic Policy Review, 8 (1), 163-178. Aoki, K., Proudman, J. and Vlieghe, G., 2004. “House Prices, Consumption, and Monetary Policy: A Financial Accelerator Approach”. Journal of Finan- cial Intermediation, 13 (4), 414-435. Arnold, L. G., 2002. Business Cycle Theory. New York: Oxford University Press. Bank of England, 2007. Financial Stability Report, No. 22. London: Bank of England. Bernanke, B. S., 1983. “Non-Monetary Effect of the Financial Crisis in the Pro- pagation of the Great Depression”. American Economic Review, 73 (3), 257- 276. Bernanke, B. S. and Blinder, A. S., 1988. “Is it Money or Credit, or Both, or Neither?”. American Economic Review, 78 (2), 435-439. Bernanke, B. S. and Blinder, A. S., 1992. “The Federal Funds Rate and the Channels of Monetary Transmission”. American Economic Review, 82 (4), 901-921. Bernanke, B. S. and Gertler, M., 1989. “Agency Costs, Net Worth, and Busi- ness Fluctuations”. American Economic Review, 79 (1), 14-31. Bernanke, B. S. and Gertler, M., 1990. “Financial Fragility and Economic Per- formance”. The Quarterly Journal of Economics, (1), 87-114. Bernanke, B. S. and Gertler, M., 1995. “Inside the Black Box: The Credit Chan- nel of Monetary policy Transmission”. Journal of Economic Perspectives, 9 (4), 27-48. Bernanke, B. S. and Lown, C. S., 1991. “The Credit Crunch”. Brooking Papers on Economic Activity, 2, 205-247. Bernanke, B. S. and Mihov, I., 1998. “Measuring Monetary Policy”. The Quar- terly Journal of Economics, 113 (3), 869-902. Bernanke, B. S., Gertler, M. and Girlchrist, S., 1996. “The fi nancial accelerator and the Flight to Quality”. The Review of Economics and Statistics, 128 (1), 1-15. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 195King, R. G. and Plosser, C. I., 1984. “Money, Credit, and Prices in Real Busines Cycle”. American Economic Review, 74 (3), 363-380. Kishan, R. P. and Opiela, T. P., 2000. “Bank Size, Bank Capital, and the Bank Lending Channel”. Journal of Money Credit and Banking, 32 (1), 121-141. Kiyotaki, N., 1998. “Credit and Business Cycles”. The Japanese Economic Re- view, 49 (1), 18-35. Kiyotaki, N. and Moore, J., 1997. “Credit Cycles”. Journal of Political Eco- nomy, 105 (2), 211-248. Magud, E. N., 2010. “Currency mismatch, openness and exchange rate regime choice”. Journal of Macroeconomics, 32 (1), 68-89. Martin, A. and Ventura, J., 2010. “Theoretical notes on bubbles and the current crisis”. NBER Working paper, No. 16399. Mody, A. and Taylor, M. P., 2004. “Financial Predictors of Real Activity and the fi nancial accelerator”. Economic Letters, 82 (2), 167-172. Mody, A., Sarno, L. and Taylor, M. P., 2007. “A cross-country fi nancial acce- lerator: Evidence from North America and Europe”. Journal of International Money and Finance, 26 (1), 149-165. Monacelli, T., 2009. “New Keynesian models, durable goods, and collateral con- straints”. Journal of Monetary Economics, 56 (2), 242-254. Myers, C. S. and Majluf, S. N., 1984. “Corporate Financing and Investment Decisions when Firms have Information that Investors do not have”. Journal of Financial Economics, 13 (2), 187-221. Nolan, C. and Thoenissen, C., 2009. “Financial shocks and the US business cycle”. Journal of Monetary Economics, 56 (4), 596-604. Olivero, P. M., 2010. “Market power in banking, countercyclical margins and the international transmission of business cycles”. Journal of International Eco- nomics, 80 (2), 292-301. Peersman, G. and Smets, F., 2005. “The Industry Effects of Monetary Policy in the Euro Area”. Economic Journal, 115 (503), 319-342. Portes, L. S. V., 2007. “Aggregate Gains of International Diversifi cation through Foreign Direct Investment: An Inquiry into the Moderation of U.S. Business Cycles”. Global Economy Journal, 7 (4), 1-36. Portes, L. S. V. and Ozenbas, D., 2009. “On Balance Sheets, Idiosyncratic Risk and Aggregate Volatility”. The B. E. Journal of Macroeconomics (topics), 9, Article 4. Spilimbergo, A. [et al.], 2009. “Fiscal Policy for the Crisis“. CESifo Forum, 10 (2), 26-32. Stiglitz, E. J. and Weiss, A., 1981. “Credit Rationing in Markets with Imperfect Information”. American Economic Review, 71 (3), 393-410. Townsend, R. M., 1979. “Optimal Contracts and Competitive Markets with Co- stly State of Verifi cation”. Journal of Economic Theory, 21 (2), 265-293. Van den Heuvel, J. S., 2002. “Does Capital Matter for Monetary Transmission?”. The Federal Reserve Bank of New York Economic Policy Review, 8 (1), 259- 266. B R U N O Ć O R IĆ: T H E F IN A N C IA L A C C E L E R A T O R E F F E C T: C O N C E P T A N D C H A L L E N G E S FIN A N C IA L TH EO RY A N D PR A C TIC E 35 (2) 171-196 (2011) 196 Van den Heuvel, J. S., 2009. “The Bank Capital Channel of Monetary Policy”. Working paper, University of Pennsylvania. Vermeulen, P., 2002. “Business Fixed Investment: Evidence of a fi nancial acce- lerator in Europe”. Oxford Bulletin of Economics and Statistics, 64 (3), 213- 231. von Heideken, V. Q., 2009. “How important are Financial Frictions in the United States and the Euro Area?”. The Scandinavian Journal of Economics, 111 (3), 567-596. Wasmer, E. and Weil, P., 2004. “The Macroeconomics of Labour and Credit Market Imperfections”. American Economic Review, 94 (4), 944-963.