The most worrying areas would appear to be excessive leverage in some high-yield bond ETFs & REITs in the mortgage market
The downfall of all investors is ultimately ?greed?. Greed can be measured not only emotionally by looking at bullish versus bearish sentiment indices but, more importantly, how much leverage investors are taking on.
The annual stressing of the US banks is meant to stave off a repeat performance of 2008 when banks had nowhere near enough capital to cushion against a hard fall. But several years on from the crisis and with the economy and housing market on the upswing, investors are more interested in the tests as a gateway for bank payouts.
The most recent tests?pulling and pushing loan portfolios and trading books to see how banks would hold up against hypothetical economic and trading shocks devised by the Federal Reserve?showed that nearly all banks would be able to endure. Only one of the 18 tested for a disaster scenario, including a peak unemployment rate of 12.1% and drop in equity prices of more than 50%, ?failed?, or its Tier-1 core ratio fell below the 5% level the Fed has set as a key measure in the test.
Seventeen out of 18 institutions passed stress tests in results announced on March 7, with only Ally Financial failing to meet minimum capital levels in the hypothetical scenario of a deep global recession and a shock to markets.
Differences in results of the Fed-run stress tests against those conducted by the banks are likely to raise questions among regulators when deciding how much excess capital they will allow the banks to return to shareholders. The Fed believes it was judging the adequacy of banks? capital plans partly based on whether the banks had ?effective? processes for estimating potential revenues, losses and capital needs. Large variations between bank-run tests and those at the Fed may call into question banks? abilities to manage risk.
The Fed
The latest rounds of the internal debate in the Fed on excessive financial risk were triggered by Fed Governor Jeremy Stein?s important speech on February 7. Stein essentially concluded that a potentially dangerous ?reach for yield? might be taking place in US credit markets, and argued further that it might become appropriate to deal with this by raising interest rates, rather than relying solely on regulatory measures.
This contradicted the longstanding Greenspan/Bernanke doctrine, which held that the interest rate weapon should only be used to control the macroeconomy. Stein won some limited support from other Federal Open Market Committee participants, and this began to concern the markets.
The Fed doves have, however, now fought back. Ben Bernanke argued last Friday that the ?reach for yield?, which is driving bond yields lower, is in part an intended result of Fed policy, and in part a natural consequence of low growth and inflation expectations. He added that, if the Fed were to tighten monetary policy too soon, then the resulting recession would actually have the perverse effect of keeping bond yields even lower, for even longer, than under current policy. But he did admit that the build-up of financial risk needs to be carefully monitored, concluding that if it becomes a problem, it should be handled by regulatory control, not interest rates.
The BIS
So how can we tell whether this really is becoming a problem? This is where the BIS research enters the picture. In the past, the key metric used by central bankers in deciding whether the economy is becoming overheated has been the output gap. When output is above ?potential?, the traditional argument is that policy should be tightened to prevent consumer price inflation from rising. The mistake made before 2008 was to place so much weight on this metric, and on the actual behaviour of the CPI, that all other evidence was ignored. In particular, falling real interest rates, surging credit growth and rocketing house prices, were given no weight in the monetary policy debate. Absolutely none!
The BIS solution is to amend standard estimates of the output gap so that they take direct account of the financial and asset price variables, notably the three just mentioned, which have been reliable signals in the past that the financial cycle in the economy is becoming over-stretched. In other words, they argue that we should not measure the output gap just by comparing real GDP with a measure of sustainable output that is derived from the real economy (i.e. labour supply, the capital stock and labour productivity). We should also take account of the behaviour of financial variables, adjusting our measure of the output gap accordingly.
Now, of course, this is a lot easier said than done. The output gap has always been a very slippery concept, even when it was applied only to the real sector. Extending this to the financial sector, and then translating the results into implications for fiscal and monetary policy, is likely to keep macro-economists busy for a long while. Therefore the results of the BIS analysis should be seen as very tentative, but Graph 3 shows the key graphs for the US economy.
The traditional measures of the output gap are shown in Panel A on the left, using the well-known OECD method based on a production function, linking output to the amount of labour and capital available in the economy. These traditional estimates completely failed in real time to identify the unsustainable economic boom that occurred from 2005-2008, and that failure contributed greatly to the policy mistakes made during the period. With the passage of time, these traditional estimates have been amended to show that a prolonged boom was indeed taking place, but the revisions came about 5 to 10 years too late to save the world from the Great Recession.
By contrast, it seems that the BIS finance-neutral output gap (Panel B) would have picked up the excessive nature of the boom much more successfully when it was actually taking place. The strong growth in credit and house prices, and the low levels of real interest rates, would all have been interpreted as signals that output was above its sustainable level, so the implied output gap was markedly positive throughout the decade before the crash in 2008. There has been no subsequent need to revise these estimates as more recent evidence has become available.
Implications for today?s debate on excess risk
The BIS method, reacting to the large declines in house prices and credit growth in the US, estimates that the current level of the output gap is about 4-5%, which is broadly in line with the latest estimates made by the CBO, IMF and OECD. Consequently, the key conclusion that I draw from the BIS work is that there has not yet been any build-up of generalised financial risk that would indicate a need to tighten overall monetary conditions in the US. This would apply either to a reversal of QE, or to a rise in interest rates.
It is possible, however, that there might be more restricted pockets of excess risk developing in certain parts of the financial system, as suggested by Jeremy Stein. The most worrying areas, in the view of the Fed, would appear to be excessive leverage in some high yield bond ETFs, and REITs in the mortgage market, where the search for yield has been particularly pronounced. It would not be surprising to see regulatory action taken in this space this year.
Over the last few years, the chase for yields, due to the Bernanke?s consistent push to suppress interest rates, has driven investors into taking on additional credit risk to increase incomes. Graph 4 is the BofA Merrill High Yield (aka Junk Bond) Index.
The graph shows that, opposed to Bernanke?s statement, investors are rapidly taking on excessive credit risk which is driving down yields. With those yields now at historic lows, there is little margin for error either in the credit markets or the economy.
Monetary action by the world?s central banks in any of the spaces outlined by me above could ultimately result in the reduction of global financial risk-taking appetite with its concomitant negative impact on world stock markets.
The author is CEO, Global Money Investor
