An extensive but benign crisis?

Tommaso Monacelli

31 August 2007



The public and (especially) the press seem to have overreacted to the current turmoil in financial markets. It is often claimed that all we are witnessing is the global-liquidity‘s revenge on Bernanke. However, if it is a financial turmoil that we are facing, it is most likely to involve an “extensive/benign” scenario rather than an “intensive/malign” scenario. An extensive/benign scenario is one in which a specific and quantitatively limited type of risk (i.e., the one related to the subprime borrowers in the US) is spread out extensively across investors and countries for risk-sharing purposes (the benign phenomenon) via the instruments of financial diversification. An intensive/malign scenario, by contrast, is one associated with a large amount of risk concentrated with some investors (possibly geographically), whose deterioration usually leads to large default losses (the malign phenomenon).

In the last twenty years, financial markets have changed dramatically throughout the world, and in the US in particular. This has been synonymous with increased ability of risk diversification. Put differently, the new financial system has become increasingly atomistic. The physical link between the primary borrower (the family seeking a mortgage) and the lender, via a plethora of instruments of financial diversification (and of subsequent borrowers/lenders along this chain), has weakened considerably.1 At the same time, technological improvements in the risk assessment process have substantially reduced monitoring costs for lenders.

In this context, the fact that lenders (loosely speaking) have been assuming an increasing amount of risk (“the subprime loans”) is a natural implication of the deepening of financial diversification. In the specifics of mortgage markets, home-ownership projects that were turned down ten years ago have now become eligible for finance. With falling monitoring costs and increased ability of diversification, financing riskier categories of borrowers can be perfectly consistent with profit maximisation by lending institutions. On the other hand, for previously constrained families, this process of financial diversification has meant a loosening of their borrowing constraints. Overall, and from the viewpoint of economic theory, it is hard to identify this as a malign phenomenon.

It is sometimes argued that, along the financial diversification chain, it may become increasingly difficult to identify where the risk exactly lies. Certainly true, yet isn’t this exactly what financial diversification is all about? Making idiosyncratic (family-specific) risk negligible relative to the aggregate pool of financed (home-ownership) projects.

From a different angle, many critics have pointed out the fallacy of this process arguing, somewhat loosely, about “excessive lending” or “excessive amount of risk” as necessary drawbacks of increased financial diversification. From the standpoint of economic theory, though, “excessive” is meaningful only if “inefficient”. In this case, one can formally identify an inefficiency if either of two phenomena arises: (i) an increased “adverse selection” and/or (ii) an increased “moral hazard” problem. Possibly, only the latter qualifies as concrete in this context.

Adverse selection and moral hazard

Isn’t it worrying that, simply allured by the rumour that “nowadays nobody is denied a mortgage”, virtually any family – including the most risky ones – decides to show up in a bank and ask for a loan? Not really, to the extent that the risk associated to this borrower is priced correctly (with this being more likely as monitoring costs fall) and is diversified through the system. After all, once again, this is what financial risk-sharing is all about.

Isn’t it true that, tempted by the increased opportunities of insurance, financial institutions have been taking up an increasing amount of risk? Prime facie, this may qualify as a deepening of a moral hazard problem. “Lending institutions need to take risks by making loans, and usually the most risky loans have the potential for making the most money. A moral hazard arises if lending institutions believe that they can make risky loans that will pay handsomely if the investment turns out well but they will not have to fully pay for losses if the investment turns out badly”.2

In the specifics of our example, the “insured” is financial institution “n-1” along the chain and the “insurer” is financial institution “n” buying a mortgage-backed security. What is crucial about moral hazard, though, is that the insured individual (better informed than the insurer about her own intentions) has the ability to affect the return distribution through her behaviour, and does that in a distorted way. Does this apply to our case? Possibly yes. Pushed by fierce competition to make it to the “funds-of-the week” top-ten list of pseudo-specialised financial reviews, with the comfortable belief that one will be handsomely compensated in the case of success and allured by the possibility of diversifying much of the risk away, many funds’ managers have probably taken up an increasingly inefficient amount of risk. A correct assessment of risk should instead consist in compensating funds managers just slightly less if the fund is listed, e.g., eleventh in the ranking (if only such an ideal ranking existed!)3. To be sure, this potential source of inefficiency does not lie in the funding of subprime loans per se, but in the excess funding of risky projects due to a perverse/distorted assessment of risk.

A correct quantitative assessment of the proportion of these inefficiently risky loans is extremely hard. However, one should make sure that such an assessment be made relative to the spectacular increase in financial investment experienced in the last ten years in both the US and global markets. In this vein, there is scope for cautious optimism.

House prices, and aggregate vs. idiosyncratic risk

In the turmoil of comments witnessed these days, many seem to have forgotten that, in the US, the initial cause of distress has been a fall in house prices. It is well-known that, via gains in home equity4, the house price acceleration has considerably widened the access to borrowing for the average family - through a series of instruments: secondary loans, mortgage-equity withdrawal, mortgage refinancing, etc. Here, though, we would like to focus the attention on two partly neglected aspects: (i) the previous increase in house prices may not have necessarily been a bubble; (ii) a fall in house prices is the realisation of an aggregate risk.

Are we really confident that the recent fall in house prices qualifies (as many have repeatedly suggested) as the pricking of a bubble? This is important; for it implies that the previous price inflation was somehow inefficient.5 However, serious models exist (the elaboration of which Governor Bernanke has eminently contributed to6) that can rationalise an acceleration in asset prices as the result of a so-called “credit cycle”: an initial increase in house prices (perfectly consistent with “fundamentals”) strengthens the demand for borrowing (via an equity valuation effect), which in turn validates and reinforces the initial increase in prices. Of course, one cannot rule out that part of the observed run-up in house prices may have been unjustified on the basis of “fundamentals”. Yet, once again, such an assessment should be made relative to the acceleration that can be rationalised on the basis of a coherent model of the type described above. Furthermore, the parallel strong acceleration in housing investment experienced in the US may have gradually led to a re-balancing of supply with demand in the housing market, finally leading to the recent fall in prices.

A possible source of concern behind the fall in house prices is that it constitutes the realisation of an aggregate shock. As it hits all families simultaneously, this shock is by definition not diversifiable. Hence, there is nothing to blame the modern financial architecture here. This is definitely material for monetary policy. Fortunately nobody knows better than Bernanke about the connections between the financial and the real side of the economy. Despite the allegations of “rooky mistake” for defining the subprime problem as “contained”, Bernanke is the one that has spoken recently about a possibly forthcoming “negative financial acceleration” problem for US families: falling house prices leading to a worsening of balance sheets, to a rise in families’ finance premia and tightened borrowing conditions, with possible final effects on consumption.7

However, this concern may once again be worth a word of caution. In today’s increasingly integrated financial markets, national (usually the prototype of aggregate) shocks assume increasingly the form of idiosyncratic shocks: country risk can in fact be shared away internationally. This entails that both the US and Europe may end up experiencing a dampening in their growth rates of consumption/output in the near future, but of possibly contained magnitude exactly because of the benefits of international risk-sharing.

The stock market and Bernanke's two sides

What to make, then, of the recent turmoil in financial markets? Here we obviously enter more risky territory. One interpretation is that the usual irrational exuberance of the market may have focused excessively on the “extensive” rather than on the “benign” part of the story. A spark originating from a somewhat limited niche of the US mortgage markets was after all spreading geographically with surprising pervasiveness. In this vein, the phenomenon was taking the form of a “new” crisis.

But couldn’t it be that we are just facing a relative benign risk being spread out extensively (and therefore not likely to generate major losses and defaults) as opposed to a malign intensive risk concentrated geographically (as the bank crises of the past, see for instance the Massachusetts credit crunch of the 1980s?).

Is the Fed hesitating too long in cutting interest rates? The malevolent interpretation is that Bernanke is hostage to his (alleged) schizophrenic identity, with the champion of inflation targeting on the one hand, and the scholar of the Great Depression on the other. More than a weakness we may see this as a strength. The Fed may well have embraced the extensive-benign interpretation. If this was the case, it is sensible to wait that the portion of “inefficient risk” (see our point above) be naturally re-absorbed by the market, thereby avoiding an ex-post validation of any moral hazard behaviour (however relevant it might have been). Different, and more important, is the issue that pertains to spillovers that may affect the real side of the economy. The Fed is definitely anticipating a cut in the funds rate if any signals of such spillover materialise. In the meantime, the international risk-sharing scenario cited above may continue to offer a comfortable buffer of inertia, both for the Fed and the ECB.





1 The IMF refers to this as a “more arm’s length” financial system, with an increased role for price signals and competition among lenders (see IMF WEO, September 2006).
2 Source: Wikipedia
3 I thank Nicola Pavoni for a lively discussion on this point.
4 Technically, the difference between the existing value of the mortgage and the market value of the house.
5 Some economists would, for goods reasons, also qualify bubbles as efficient outcomes, but we prefer to abstract from this point here.
6 Bernanke and Gerler (1989), Kyotaki and Moore (1997). These papers differ in many details but both contain a financial acceleration mechanism.
7 See





Topics:  Financial markets

Tags:  house prices, subprime crisis, Bernanke, crisis, family economics

Professor of Economics at Università Bocconi, Milan and CEPR Research Fellow