Attend Spring Meetings on Development topics from Apr 17-21. Comment and engage with experts. Calendar of Events

Global Economic Prospects 2015

The global economy is still struggling to gain momentum. Read more ...

Development Prospects

Providing information, analysis, and advice on global trends in the world economy.  Find out more ...

Global Economic Monitor

Now free. Daily and monthly updates on global economic developments and relevant topical issues. Find out more ...

Syndicate content

How Exactly was the Fed Responsible for the Financial Crisis?

Jamus Lim's picture

As Alan Greenspan has famously insisted---and recently reiterated (PDF)---the Fed's culpability in the entire credit bubble is limited to its failure in its (secondary) role as regulator, and not in its (primary) role as monetary authority. Given how the Fed only controls the short-term rate, the argument is that since the close relationship between short and longer-term rates broke down in the 1990s, one can hardly blame the Fed for engineering a bubble with the tools that it has at its disposal. Ergo, even with clear Taylor rule deviations, a disconnect still exists between Fed policy and the bubble.

Some authors have taken strong exception to the claim. Notwithstanding the fact that short and long run rates do not move synchronously, it is nevertheless interesting to query the precise transmission mechanism in short rates may have affected home financing, through either long rates or otherwise, in the most recent crisis episode.

First, one needs to ascertain whether a disconnect truly does exist between long and short-term rates. We use 30-year Treasuries and the median conventional fixed mortgage rates for the former, and the effective Fed funds rate for the latter. From the figure, it is clear that the three series track each other closely for much of the 1980s. However, there appears to be some erosion in this relationship starting from recession of the early 90s, before returning to the tighter pattern (albeit with narrower spreads) in the second half of the decade. The series once again diverge in the 2000s, especially in the post-dot-com recession period, where short rates appear to have little influence on longer rates.

This point can be made (very slightly) more formally, although the results are a little more surprising (see table). Taking simple period correlations, we see that the correlation between the Fed funds rate and 30-year Treasuries clearly fell from when comparing the 1980s to 1990s. The relationship between the Fed funds and mortgage rates also fell, but the relationship remains somewhat stronger than with long bonds. But these correlations pick up in the first decade of the 20th century. What is most striking here is that the correlation between the Fed funds and fixed mortgage rates rises to a value almost as high as in the 1980s. So while there may be something to the claim that there was a disconnect between short and long rates more generally, this was certainly not the case for the Fed Funds rate and fixed mortgage rates. This distinction---between mortgage rates and long bond returns---hints at the second issue.

   Corr(Fed funds, 30s) Corr(Fed funds,mortgage)
 1980s  0.69 0.79 
 1990s  0.37  0.57
 2000s  0.51 0.74 

Source: World Bank staff calculations, from Datastream

Second, financial markets---or more precisely, mortgage-related financial instruments---were an important mechanism this time round that transformed the low Fed funds rate into cheap mortgage financing, which in turn fed the housing boom. While mortgage-related securitization had been around for a while (mortgage backed securities were pioneered in 1970 by Ginnie Mae, and CMOs and CDOs were brought to the market in the mid-1980s), widespread mortgage-related securitization (and resecuritization) only took off in the 2000s. The (then) stellar ratings issued on these instruments meant that they became attractive investments, and financial market participants of all shape and sizes borrowed from short-term markets (which recall is directly influenced by the Fed) to fund these purchases.

Moreover, the popularity of such instruments---along with their perceived safety---also meant that MBSs and CMOs and CDOs were often posted as collateral for repo transactions. These repo transactions, in turn, encouraged rehypothecation, which allowed aggressive firms to massively increase their leverage.

Finally, add to the mix how option ARMs---which are principally concerned with the prevailing short rather than long term rates---had become increasingly popular as a means of financing (and whose nonperformance predated the subprime mess), and you have multiple financial innovations that worked to short-circuit (literally and figuratively) the traditional short/long transmission mechanisms and inflate a housing bubble.

A third and final point. Although new financial instruments appear to have played an important bridesmaid's role in the recent crisis, there are in fact more standard ways that the Fed funds rate can affect long rates and economic activity. How? This calls for a brief theoretical digression.

Theories of the term structure of interest rates, especially those concerned with the shape of the yield curve, offer some insight into the manner by which short rates may impact longer ones. While several possible candidate explanations exist, most would agree that a key factor is the strength of market expectations concerning the future path of interest rates. Furthermore, the monetary policy transmission mechanism has been shown to operate through the channels of both credit as well as money.

With these ideas in hand, we can see how, even with short and long rates out of (empirical) sync, actions by the Fed can nonetheless affect expectations, and such expectations (of continued cheap money) may be sufficient to drive excessive mortgage lending. Furthermore, a low Fed funds rate can instigate banks to seek avenues for expanding short-term credit: In particular, these seem to have manifested themselves in lax lending standards.

What lesson does this somewhat esoteric discussion hold for developing economies? Since most developing countries do not have the same deep, liquid, and mature financial markets as the United States, they are more insulated from possible excesses generated by complex financial instruments. Still, many emerging markets actually tie their mortgage rates to variable benchmarks, and credit standards may actually be weaker in countries with generally lower levels of institutional governance. As developing country central banks ponder the appropriate exit strategy from their stimulative monetary policies effected in response to the global crisis, it would be helpful to keep in mind how extended monetary ease, even if the cuts were originally appropriate (the Greenspan Fed slashed rates in an environment that included the bursting of the dot-com bubble and the September 11 attacks, after all), there is a need to be vigilant about when they may cease to be so.


Submitted by Anonymous on
Since approximately 80% of U.S. mortgages issued in recent years to subprime borrowers were adjustable-rate mortgages ARMs I am not sure what this has to do with either 30 years Treasuries or fixed rate mortages. Then it says: “mortgage-related financial instruments---were an important mechanism this time round that transformed the low Fed funds rate into cheap mortgage financing, which in turn fed the housing boom.” Though this is partially true let us not forget that what really brought on the crisis was the transformation of bad risks into supposedly low risks, by means of too favorable credit ratings, in natural response to the extraordinary regulatory incentives awarded to anything connected with good credit ratings, by means of ridiculously low capital requirements for banks.

Submitted by Jamus Lim on
Thanks for the comment. In response, I'll just note two things to your two observations: (1) It is true that a large number of mortgages issued in the subprime market were ARMs (and hence tied more to short-term rather than long-term rates), and that the crisis is often regarded to have originated in the subprime sector (although this has been disputed, but that is another story). Still, the claim I was dissecting was the one made by the Fed (not by me) that their role in the housing bubble was limited because long rates had become untied from short rates. Moreover, saying that the bubble has nothing to do with long rates because subprime mortgages were mostly ARMs implicitly assumes that the subprime sector (which is between 10--20 percent of the entire housing market) was the only part of the housing market in which there was a bubble. While this may be true (bubbles are after all very difficult to identify), I am nonetheless very hesitant to make this case. (2) I do not deny that the factors driving the crisis were multifacted, and the mispricing of risk (with the rating agencies' blessings) were undoubtedly a factor. In fact, this is precisely the point of my second point (no pun intended). But poor risk pricing is at best a proximate determinant. We need to identify the more fundamental ones if we are to learn from the crisis, and these are likely to be excessively easy monetary policy (as I discuss in the post) or regulatory failure (as you and many others, including the Fed itself, have pointed out) or both. The post was trying to clarify the channels by which the former determinant would operate.

Add new comment