1 Introduction

We are besieged, daily, with reports of corporate collapse, the demise of credit markets and the vulnerability of our banking system. Yet, the explanations we receive are usually in the form of statements as to the “complexity” of “exotic” financial products, the “sophisticated” risk management tools used by the banking system, and the globalisation of financial markets. One has to wonder what sort of risk management has been in practice, what sort of economic theory has been in the minds of regulators, and what kind of economic and financial leadership has led to such dire circumstances. In fact, the circumstances, we are told, are so dangerous, that only the immediate injection of trillions of dollars will be able to mitigate the approaching debacle. The cash must be given, they say, to the banks without delay and without oversight or else we risk an economic depression, at the very least, as serious as that experienced in the 1930s. What has happened in the last decade?

2 Some Trends

Let us first consider some trends that have been visible for a while. Shiller (2007) notes that although it is popular to explain the recent boom in housing prices as being a result of low interest rates, the rise did not actually begin until the mid-1990s, long a fter nominal interest rates had retreated from their highs in the 1980s. Figures 1 and 2 (p 138) illustrate this phenomenon.

As stated by many economists, the low interest rates led to borrowing booms and an enormous debt build-up. According to Hudson (2006), American households are now deeper in debt than at any point in history. In fact, mortgage loans now constitute close to 90% of the increase in debt since the 1990s and make up fully 50% of bank loans in general. This trend can be seen to be driven by a combination of factors, including record low interest rates, which increase the borrowing capability of home buyers,

STRUCTURAL CAUSES

favourable tax treatment of mortgage interest, and the “wealth effect” (the increased spending caused by the recognition of the value of one’s home) benefits to the general economy. In Figures 1, 2 and 3, we see a number of other important trends.

Figure 1: Long-Term Nominal Interest Rates(4 countries)*

16

India 14

UK 12 10 8 6 4 Euro

2 Japan

0 1953 1957 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005 2009 US

* Source: Global Financial Data.

Figure 2: United States House Prices($)* 300,000

250,000

200,000 Inflation-adjusted house prices

150,000

100,000

50,000 Nominal house prices

0 1987 1989 1991 1993 1995 1997 1999 2001 2003 2004

20062007

* Source:http://housingbubble.jparsons.net

Figure 3: Real (inflation-corrected) House Prices (7 countries)*

UKUK

ance

FrFr

Australia

Australia

USUS

Italy

Italy

German

yy

German

Japan

Japan

1987 1989 1993 1997 2001 2005

1991 1995 1999 2003

2006

* Source: OECD Economic Outlook 2007, Statistical Annex.

First, there is the increase in the real price of housing, for example, as measured by the Case-Shiller index. This is illustrated in Figure 2.

With the exceptions of Japan and Germany, we see a similar trend in the UK, France, Australia, Italy, and (again) in the US. Organisation for Economic Cooperation and Development (OECD) data for these countries is presented in Figure 3.

Again, we see the phenomena from another point of view, sales volume, in the US, in Figure 4. The data are quite dramatic.

As seen in Figure 3, the phenomenon is not limited to the US. In fact, many other countries have experienced a similar upward

138 trend in housing prices. In fact, Ayuso and Restoy (2006) have estimated that Spanish housing prices were overvalued by as much as 32% as of 2004. In Figure 5, they have illustrated the growth (Spanish) in house prices, rents, and the price-to- rent ratio.

A similar trend is noted for the US in Figure 12 (p 143). Although their predictive power seems clear, Andrea Finicelli (2007) has criticised the use of the price-to-rent criterion. She points out that it is not enough simply to look at prices with respect to rents because even if prices are justified, they are simply not affordable to consumers. In a market such as housing, this is an important point. The charts illustrate factors connecting the present value debt payments and present value income for typical home purchases in the US. We note that until 2003, the affordability of homes, as measured by this ratio, stayed highly correlated with interest rates. After 2003, we start to see a strong divergence. The affordability issue is particularly critical when consumers face credit constraints. In this case, sensitivity to interest rates increases.

Figure 4: Existing 1-Family Home Sales* 6400

6200

6000

5800

Existing 1-family home sales

5600

5400

5200 5 Per Mov Aug (Existing 1-family home sales)

5000

4800

4600

4400 1999 2005

* Source: National Association of Realtors/Haver Analytics.

Figure 5: Real House Prices, Rents and, Ratio, q, Price/Rent* 300

Real housing prices200 Real rents

q

100 1987 1989 1991 1993 1995 1997 1999 2001 2003

2004

* Source: Ayuso and Restoy (2006).

Figure 6: Index of Housing Affordability and Nominal Mortgage Interest Rate*

Affordability Index

Interest rate

1975 1978 1981 1984 1987 1990 1993 1996 1999 2002

2005

2006

* Source: Federal Reserve & Office of Federal Housing Enterprise; the index is calculated by multiplying the Housing Price Index by the Mortgage Interest Rate and dividing by Per Capita Disposable Income.

march 28, 2009 vol xliv no 13 EPW Economic & Political Weekly

STRUCTURAL CAUSES

Figure 7: Debt Service Ratio*

15

14

13

12

11

10 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2006

* Federal Reserve; the ratio is of the sum of interest and minimum contracted principle to disposable income.

Figure 8: US Mortgage Delinquencies

4.0 3.6 3.2 3.0 2.8 2.6 20 18 16 14 12 Prime Segment Subprime Segment

10

.

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Source: Mortgage Banking Association.

Figure 9: Rated CDO Volume ($ million)* 300,000

250,000

200,000

150,000

100,000

50,000 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Rated Tranches of CBO/CLO

* Source: Moody’s Investors Service.

The increasing indebtedness of the American household is shown in Figure 7; here we depict the debt service-to-disposable income ratio. As shown, the ratio went up to an unprecedented high.

We would like to notice other trends, ostensibly independent of interest rate trends. These are, first, the alarming rise in mortgage delinquencies; second, the growth of Collateralised Debt Obligations (CDO) during the same period – substantially contributing to the US sub-prime crisis. This is shown in Figure 9.

Figures 8 and 9 will play an important role in our modelling of the outbreak of the financial crisis (see Sections 5 and 6).

A combination of hysterical warnings and incidents, involving individuals and institutions we were accustomed to trusting, such as Bernard Madoff and Lehman Brothers, combined with the seemingly incomprehensibility of the issues involved led e veryone from politicians to journalists to join the clamour to a llocate huge sums to very risky assets, for example, credit derivatives. Yet, the issues were not really as complex as one might b elieve. This does not mean the problems are easy to fix, but far more people are quite capable of understanding the underlying phenomena than might be expected.

Now, although it is certainly true that the methods used in the pricing of credit derivatives, the computation of value-at-risk,

Economic & Political Weekly

EPW

and in the modelling of structured financial products involve such concepts as measure theory, Ito’s Lemma, and jumpdiffusion processes, it is not necessary to understand them in

o rder to have a basic grasp of the current situation in the financial marketplace. Further, as has been underscored by such writers as Nasim Taleb, these methodologies are of dubious value.

3 Some Explanations

Next, we want to briefly review some theoretical explanations of the outbreak of the financial crisis in the housing sector. In most of the recent models, the role of the dividend-price ratio, the rent-price ratio, is seen as a crucial variable. Specifically, concerning housing as an asset, there seems to be little debate that interest rates and the dividend-to-price ratio were critical in determining the growth of the market since the 1990s. Campbell and Shiller (1988) used the intertemporal pricing equation

⎡ k ⎛ 1 ⎞i ⎤⎛ 1 ⎞k

Pt = Et ⎢∑⎜ ⎟ dt+i ⎥ + Et ⎜⎟ Pt+k

⎢T=1 ⎝ 1 + δ⎠⎥ ⎝ 1 + δ⎠

⎣⎦

where δ is the discount rate, P is the price, and d is the dividend (rent, in our case). They showed that, after taking logs, one could derive an expression for the rent-to-price ratio

∞

dt − pt = Et ∑⎛⎜ 1 ⎞⎟ t ((rt+i – d^) t=1 ⎝ 1 + δ⎠

where d^ is the growth rate of the rent and r the asset return. The

t

above formula means that if, for example, the growth rate of dividends is zero, the dividend-to-price ratio is determined by the discounted expected returns. This determines the rate at which the asset is consumed. The ratio, in turn, can lead to a predictor variable as follows:

= a + b (d – p) + εt+1

rt+1tt

Here, the rent-to-price ratio is seen to be a good predictor for the return on real estate assets. Many attempts have been made to explain the run-up in housing prices from a macroeconomic perspective. Mostly, they have focused on considerations of the future income that real estate can generate as compared with current prices. More specifically, the dominant theories revolve around varying interest, growth, and discount rates. For example, in Ayuso and Restoy’s work, a model of the net return on a house for a single period is given thus:

1

r = d + q − q − k

t+1t+1t+1t

δ

where d is the real growth rate of rents and q is the price-to-rent ratio and δ and k are linearisation parameters. Assuming a constant risk premium (an issue we will come back to later), the equilibrium ratio, q*, is

δ∞ s q *t = (k −πm) + Et ∑δ [dt+s − rm,t+s] 1 −δ s=1

where π is the difference between r, as defined above, and the r eturn on some reference portfolio, m.

STRUCTURAL CAUSES

An interesting point to note here is that, under normal circumstances, the second term in the intertemporal equation is usually negligible, that is, for large k and reasonable δ. However, in a boom, fast turnover is common, thus the relative importance of the rent and/or any ratios that follow from it, may be ignored. In other words, speculators plan on a quick resale and disregard the consistent cash flow.

There are many macro models that take the price-to-rent ratio as a starting point. However, these models tie macro variables to micro variables. What is needed is to explore how an individual decision-maker, operating with knowledge of these same macro variables would act. In other words, we intend to examine the microstructure of the housing market.

On the micro level, there are intricate processes operating s imultaneously; these, in turn, introduce more complexities than any of the macro models discussed so far have taken into account. On the other hand, the interplay of these processes is easy to u nderstand and points to simple conclusions. We hypothesise a micro-structure that will explain not only the run up in housing costs, but also the subsequent collapse and ensuing credit crisis.

Further, we will show that US federal policy is relatively weak in controlling housing booms and busts and that issues of asymmetric information are far more relevant than recognised. A dditionally, our model will demonstrate how underlying market instruments can influence prices and macro-phenomena in e xtreme ways.

We also examine the role of correlation in the default process. On closer examination, we will see that this is not only not s urprising, but also essential. The short explanation is that the housing market, unlike many markets, is highly dependent on the banking system. Because most homes are purchased through financing, the underlying structures which allow for the allocation of risk and funding are critical factors in pricing. Correlation will prove a critical component in our analysis. Yet, before we get to this, we have to discuss banking operations.

4 The Crucial Role of the Banking System

We would like to understand how things fit together by taking the banking system into account. We begin by identifying the key components of bank operations. Basically, banks collect savings and lend it to customers who want to borrow. Borrowers, for e xample, households, firms, investment banks, or hedge funds, in need of cash get it from the bank. In return the borrower pays the bank a larger rate on the sum. The bank’s profit is the difference between the two rates. So far, so good.

However, there are several potential difficulties. Let us suppose that the funds have been borrowed by a household for the purchase of a home. In this case, the bank has several worries. First, what happens if the borrower does not pay his loan? This is called default; the obvious remedy is that in such a case, the bank gets to keep the property purchased. This is foreclosure. The bank takes the property and sells it. Unfortunately, seizing someone’s property is a complicated legal issue. This implies costs. Thus, the bank will not collect the full value of the loan. This is called the recovery rate and is also expressed as a percentage. A recovery rate of 75% implies that in the event of default, the bank

140 will only be able to recoup 75% of the amount loaned. This is likely to happen more as delinquency rates rise, as shown in Figure 8. The risk that the bank will not receive what it is owed due to the extension of credit is the credit risk.

The second worry is that the borrower will find a cheaper source of financing. In this case, the borrower will borrow at the cheaper rate, pay off the original loan, and leave the bank with no income to pay its depositors. In this case, the bank still owes its depositors the agreed upon rate, but has no source of cash to pay it. This risk that the bank will not be able to meet its obligations due to fluctuations in income is called interest rate risk.

The third worry is that regulators, who are aware of the potential problems faced by banks, insist that banks keep, in cash, a certain percentage of the cash they have loaned out. This is called the liquidity problem and refers to regulatory capital. The problem for the bank is that this constitutes liquid capital which is, in some sense, frozen, that is, it cannot be loaned out, and thus generates no income. Banks would like to lower their risk profile, thus freeing up this capital for loans, which they often do using clever accounting.

It is clear that banks would probably be interested in any scheme that would allow them to avoid these three pitfalls. Further, the sudden growth in cash flows from commodity-rich, and other emerging economies generated a lot of demand for suitable products. Usually, people want to invest their cash in a secure place that earns a high rate of return. As shown above, during the last decade, bank and government interest rates have been notably low. By contrast, in the 1980s, one could achieve double-digit returns in one’s local savings bank. Thus, the last couple of d ecades have been characterised by large quantities of high r eturn-seeking investments.

Yet, there was the dotcom bust which became apparent after the 9/11 attack. Huge sums had been invested throughout the 1990s, similar to the technology bubble of the late 1920s. The collapse of this market scared investors. However, like a good surfer, when one wave collapses, one does not go home. Instead, one finds another wave. So, the next economic catastrophe was brewing. Along with the hi-tech revolution of the 1990s, came the spinoff technologies that made global finance possible. The a dvent of the computer and the Internet made it possible for huge sums of money to be moved around the world in seconds, for i nvestors in one continent to participate in markets in another continent, and for complex computations, back-office, settlement, and clearing house operations to be carried out at unheard of speeds. The growth of the Bloomberg and Reuters business i nformation networks made vast amounts of data available as never before.

In our own work (Grüne, Semmler and Bernard 2007-08) we have shown the dangers of having all one’s “eggs in one basket”. Yet, the world’s largest financial institutions seem to have lost sight of this fact. In that paper, we focused on evaluating a company’s capital assets and credit risk in the context of a productionoriented asset pricing model. Interestingly, some preliminary thoughts on the relationship of credit and a firm’s capital assets can be found as far back as in Keynes (1936). Using this formulation, we considered the evaluation of the default risk of a company by solving a debt control problem treated as a dynamic decision

march 28, 2009 vol xliv no 13

STRUCTURAL CAUSES

problem. This solution, in turn, led to the company’s asset value. We then extended this method to include multiple capital assets subject to random shocks. In this way, we examined how the default probability is related to the correlation of shocks to the different fractions of a company’s capital assets.

5 Complex Securities

Having all the peripherals in place, we can start to envision the sort of structures that will connect the central sections. We start with the idea of a securitisation product. Contrary to what one hears on television, anyone with a rudimentary knowledge of probability is fully capable of understanding how these products work. For the time being, we will use a simple example to explain the mechanics. Specifically, imagine flipping a coin. Each side is equally likely to appear. Thus, we say that Probability (heads) = 1/2. Also, each time we flip the coin, the chances of getting “h” are exactly what they were before. In other words, the probabilities remain fixed. Further, each time we flip the coin, it has no effect on subsequent outcomes. In other words, the probabilities are independent of each other. Let us now imagine that we flip the coin four times and record the results. For example, one such sequence might be {h,h,t,h}. Again, there is nothing special about this particular string. In general, the probability of {t,t,t,t} is e xactly the same as that of {t,t,t,h}. Why is that? The answer can be seen in the following equation:

n

Pr (X = k)= pk (1– p)n–k

(k)

Here, n = 4, k = number of heads in four flips, and p = ½

So, it is easy to realise that while the chances of getting one head in a single flip of a fair coin is ½, the chances of getting one head in four flips of a fair coin is only 4/16 = ¼. This is the “magic” that is behind securitisation projects.

To understand this, let us consider mortgages instead of coins. Suppose we have four mortgages instead of four coins. Let us also suppose that the mortgages are very risky and that the chance that any one of them will default is, like the coin, ½. Additionally, we suppose that they are in different cities, so there is no relationship between them. Mathematically, the situation is identical to that of the coins.

Now, each mortgage is supposed to pay some interest. Thus, there is a positive cash flow from this collection. If one could find someone willing to guarantee these mortgages, there is a source of funds with which to pay them. Such an individual might be u nwilling to guarantee an individual mortgage because it is too risky. However, that same person might accept the wager if instead of guaranteeing a single mortgage, he/she only guarantees exactly one default in a group of four, thus cutting the risk in half from ½ to ¼.

Of course, as those familiar with these products will argue, the situation is vastly more complicated. Indeed, they are correct; but it is not more complicated in the way that turbulent fluid flow is complicated.

What is important is to realise that by gathering a collection of risky assets together, one can reassign the risk into different classes, called tranches. In our example, the tranches would

Economic & Political Weekly

EPW

c onsist of one, two, three or four assets defaulting. We have already seen that the risk of exactly one asset defaulting is ¼. A clever trader could take a long position in this risk tranche and go short in another tranche. Also, in real life, the chances of default are usually supposed to be much lower than the 50:50 example we used here. There are many variations. For example, the different assets might not, in fact, be independent of each other. In other words, one asset defaulting makes the chances of others defaulting more or less likely. This phenomenon is called correlation and, along with default risk, is an important driver of the overall structure. Further complications can be introduced by mixing different risky assets of different types, taking more complex positions within the tranches, and by investing in multiple p roducts. All this is usually analysed on computers, which keep track of the details.

To put it another way, a Mortgage Backed Security (MBS) is a type of CDO in which the defaultable assets are mortgages i nstead of bonds or Credit Default Swaps. We note that the rise of this industry has exactly mirrored the housing boom as shown in the figure earlier. These securities are simply more implementations of the scheme described above. MBSs operate by grouping together mortgages and using the interest income thus produced to compensate investors for taking positions in which varying levels of default, called tranches, are guaranteed. The incentive to form such a structure is motivated primarily by the surplus cash generated – that is not needed to compensate investors. The monthly interest payments from those mortgages are income to the Special Purpose Vehicle (SPV), a company designed for this purpose. Different tranches are assigned with appropriate attachment points. If the number of defaults remains below the lower attachment point, the investors in that level simply collect the pre-arranged premium. However, once the percentage exceeds the lower attachment point, defaults are paid out of the capital posted by the investors of that tranche. Once the upper attachment point is reached, the next tranche takes over since the lower tranche is effectively exhausted. Investors in the MBS will demand compensatory interest commensurate with the assumed default risks and recovery values. These are paid from the interest income from the mortgages. The difference between the two cash flows is profit to the SPV investors. As long as it is profitable to construct these instruments, liquidity in the mortgage market will only be limited by the default probabilities, the recovery values, and the rates obtainable elsewhere. Interested readers can find out more about this industry in T avakoli (2003).

6 The Risk-Taking Investor, the Outsourcing of Risk, and the Collapse

At this point, some readers may ask: “Why would anyone take a position in such an insane structure?”. The answer is that investors are fully aware of the risk they are taking. They weigh their losses by the chances of that loss actually occurring. So, for example, the loss generated by four defaults is discounted by 1/16 because these are the chances of this actually taking place. An investor would demand that someone pay him/her 16 cents for each dollar he risks because that is what he expects to lose. To put

Figure 10: S&P/Case-Shiller Home Price Indices

10-City Composite 20-City Composite

STRUCTURAL CAUSES

V

2 t+σ

V

vestors, define the tranches, etc. As long as all the assumptions

r

−

Vt = V0e

24

20

a simple MBS structure with five reference assets (mortgages) and

16

16

three tranches, equity, mezzanine, and senior, a bsorbing the bot

12

12

tom 10%, the second 10%, and the last 80% of the default risk,

8

8

% change, year ago

4

respectively.

0

-4

-4 Figure 11: Monte Carlo Results

-8

-8

4.00

-12 -12

-16	-16
-20 1988 1990 1992 1994 1996 1998 2000 2002	-20 2004 2006 2008
Source: Standard and Poor’s and Fiserv.

Clearly, in the current crisis, a key assumption that was i gnored, often deliberately, was the default rate (see Figure 8). MBSS were constructed using historical d efault rates. These rates are not accurate for a variety of reasons. First, with the introduction of MBSs, bank no longer held the risk. Thus, the incentive for accurately assessing a potential borrower’s reliability was removed since the bank would divest itself of the mortgage within a few months of its being issued. Second, the advent of it another way, he/she expects that for every 100 positions he takes, 16 of those defaults will result in the loss of his/her dollar, but that in 84 of those positions, his/her money will simply be returned. Thus, the fair return on his/her investment is 16% – quite nice sounding in a financial environment where banks r eturn 1% on savings accounts.

3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 Housing SPV Profit Recovery rate Default rate

Now, of course, the reader may protest, saying: “But if he/she expects to lose 16%, this return only gives him/her a reasonable probability of breaking even.” There are several “solutions”. The investor may be offered 18%, for example. Alternatively, he/she may be allowed to participate with little or no money down – a highly leveraged situation. Cynically, some investors may not even realise the true nature of their bet. For them, 16% just sounds great. The point, however, is that it is easy to imagine that there are ways to use these instruments to reallocate the costs of default by reallocating the stream of cash that comes from the risky assets, mortgage payments, in our case.

The secret is to figure out how to estimate the risk of default and use this to calculate a “fair” price. In Schönbucher (2003), a large number of methods are surveyed.

We are now in a position to assemble the pieces. Banks wish to discharge their risk because it frees up regulatory capital and removes interest rate and credit risk from their books. Investors seek high-margin, high-return investments. Pension funds see the higher-level tranches as low risk opportunities. Hedge funds gain an opportunity to diversify into areas not connected with their current portfolio.

As we mentioned earlier, special companies are created, called SPVs which gather the risky assets, transfer the cash flow to infirst two factors spurred the real estate market, speculators e ntered the markets with the intention of flipping properties. Fourth, in a rising market, the possibility of foreclosure was less scary for those who ended up “holding the bag”, as it were. Lastly, and cynically, many participants in the industry found they could make easy money by simply writing mortgages to almost anybody.

We note, as is well known, a MBS is highly sensitive to default rates. A change in only a few per cent can turn a SPV from a profitmaking proposition into a losing one. When interest rates adjusted upwards and people found that they could not pay the debt service as easily as they had expected, default rates rose. Changes in the correlation between risky assets turned previously profitable positions into losers. With investors less willing to guarantee risk, banks became less willing to lend, because they had nowhere to unload the debt. Now, they are a ttempting to hold on to liquidity. Thus, just as we had a positive feedback loop on the way up, we now have a negative feedback loop on the way down. We read the results in the newspapers on a daily basis.

The best way to model this type of complex system is with a Monte Carlo simulation. In this methodology, a computer model is built and then allowed to run thousands of times, each with different random inputs. Thus, the total picture emerges.

We have adapted a method of CDO pricing, see (Chacko et al 2006) consistent with our model. Specifically, we use Merton’s (1974) definition of default wherein default occurs when the size of the debt exceeds the value of the asset. In Merton’s model, a sset valuation may be expressed as follows:

⎛⎜⎜⎝

2

0 1 2 3 4 5 6 7 8 9 10

adjustable rate mortgages (ARMs) and other, more exotic packages, allowed potential borrowers to d elude themselves into These results are hardly surprising. We see certain aspects of believing that they could handle the debt service. Third, as the the US real estate market captured.

march 28, 2009 vol xliv no 13

EPW

STRUCTURAL CAUSES

Figure 12: Home Prices/Rents	By January 2006, the risk-free rate had risen to 4.31% while
15	the ratio of housing prices to rents declined to only 13. Thus, the
	implied risk premium was actually negative.
	7 Conclusions
	The answers are not to be found in a new set of regulations of
	the type exemplified by Basel II. Such systems only spawn a
	new industry of regulatory arbitrage in which clever folks
	figure out ways to defeat them. Rather, we believe that new
	rules require institutions to retain some of the risk that they
	have generated. For example, one such requirement might
	be to insist on high pre-payment penalties for mortgages,
1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Source: Federal Reserve, Bureau of Economic Analysis.	thus discouraging refinancing, and removing some of the inter
	est rate risk from banks. Another rule might make it criminal
What is interesting in Figure 12 is that we can see that for the	to certify someone as creditworthy when they clearly are
period 1955 to 2000, the ratio stayed close to 10. This makes sense	not. Certified Public Accountants already live under such
from an investor’s point of view if the discount rate (the sum of	rules; why not mortgage brokers as well. There are many
the risk-free rate, the inflation rate, and the risk premium)	other ideas that are worth trying and we do not disparage
r emained close to 10%. However, the sudden change to 14 implies	them by not mentioning them here. Our prescription is
a discount rate of less than 6% (5.84%). Considering that the risk	basically this: if we want to continue to have private profitabil
free rate in 2005 was over 2% (averaging close to 2.25%) and that	ity, it must come along with private responsibility; we do not
inflation was averaging close to 3.39%, this means that the	advise bailouts for the irresponsible, the greedy, and the cor
i mplied risk premium was 0.20%. In other words, whereas the	rupt; if an institution is “too big to fail” and is rescued by public
traditional risk premium on housing was close to 4%, in the hous	funds, an appropriate oversight board is needed to oversee the
ing bubble, investors simply shrugged off risk altogether.	flow of those funds.

14 13 12 11 10 9 8 Average 1960 to 1995 = 10.0

References

Ayuso, J and F Restoy (2006): “House Prices and Rents in Spain: Does the Discount Factor Matter?”, D ocumentos de Trabajo #9, Banco de Espana.

Campbell, J and R Shiller (1988): “Stock Prices, Earnings and Expected Dividends”, Journal of Finance, American Finance Association, Vol 43(3) pp 66176, July.

Chacko, G, A Sjöman, H Motohashi and V Dessain (2006): Credit Derivatives: A Primer on Credit Risk, Modelling, and Instruments, Upper Saddle River (New Jersey: Wharton School Publishing).

Finicelli, A (2007): “House Price Developments and Fundamentals in the United States”, Bank of Italy Occasional Paper No 7.

Grüne, L, W Semmler and L Bernard (2007-08): “Firm Value, Diversified Capital Assets and Credit Risk: Towards a Theory of Default Correlation”, Journal of Credit Risk, 3(4), Winter.

Hudson, N (2006): “The New Road to Serfdom”, Harpers Magazine, May.

Keynes, J M (1936): The General Theory of Employment, Interest and Money (London: Macmillan).

Merton, R C (1974): “On the Pricing of Corporate Debt: The Risk Structure of Interest Rates”, Journal of Finance, 2: 449-70.

Economic & Political Weekly

EPW