Markets trapped between euphoria and despair
“Don’t panic!” was good advice provided by Lance-Corporal Jones to his commanding officer in the 1970s BBC comedy series “Dad’s Army”. Perhaps it should now be directed to central banks and increasingly jittery investors.
The last six months have witnessed a rollercoaster as markets and policymakers have alternated between euphoric optimism and crashing pessimism with bewildering speed.
Many seem convinced the world’s major economies are poised on either the brink of liquidity-induced inflation; a renewed descent into recession and deflation; or perhaps both at different times, with near-term disinflation is followed by an upsurge in inflation later.
But what if policymakers and investors are overestimating the likelihood of extreme outcomes? Past experience suggests outcomes are much more likely to fall somewhere in the middle, as the economy “muddles through”; the likelihood of extreme outcomes being realised is actually very small.
U-SHAPED FORECAST DISTRIBUTIONS
The best example of the increasingly binary outlook shared by policymakers and investors is a chart buried in the Bank of England’s May “Inflation Report” (Chart 5.13) which shows the Bank thinks it is much more likely inflation in 2012 and 2013 will be very low (<1.5 percent) or very high (>2.5 percent) than a moderate level near the target (1.5-2.5 percent).
Rather than a bell-curve shaped distribution (in which expected outcomes cluster around the middle) the Bank’s inflation forecast is U-shaped, with extreme outcomes dominating more moderate ones.
The Inflation Report focuses specifically on the UK economy. But the same binary, apocalyptic thinking is common across a wide range of asset classes and geographical markets. In its recent medium-term outlook, the International Energy Agency characterised the oil market as stuck in an “uneasy truce” between fears about supply-side tightness in the medium term and concern over a renewed deterioration in demand if global growth falters.
Investors and policymakers seem convinced the outlook is dominated by two different outcomes — one very bad and the other very good — with little in between.
With the global recovery poised on a knife-edge, even a small shift in sentiment or more fundamental factors shifts the economy from one track to the other and back again, creating equivalent gyrations in commodity and asset prices.
FOCUSING TOO MUCH ON SMALL RISKS
Evidence suggests decision-makers and investors often fail to analyse extreme outcomes properly.
In a famous 1981 paper on “The framing of decisions and the psychology of choice”, Nobel Prize-winner Daniel Kahneman and Amos Tversky, founders of behavioural economics, showed they assign too much weight to high-consequence outcomes that have a low probability of occurring (a flu pandemic or economic depression), while simultaneously underweighting middling outcomes with a much higher chance of being realised.
There are indications this may be happening at the moment as investors and policymakers assign too much weight to extreme outcomes (high inflation or outright deflation) at the expense of more moderate ones (steady rates of fairly low price increases).
In a speech late last month, Adam Posen, an external member of the Bank of England’s Monetary Policy Committee, admitted UK inflation has crept higher as the Bank has kept rates ultra low to ward off the small but not negligible possibility of a “terrible downside risk”.
When that downside risk failed to materialise, inflation gradually rose. Higher inflation was the price worth paying to avert a small but catastrophic outcome, in Posen’s view.
We will never know whether downside risk failed to materialise because policy was successful in averting it, or because the Bank over-estimated its likelihood in the first place. Even if inflation creeps higher after the fact (ex post), it does not mean the decision to keep interest rates ultra-low was the wrong decision beforehand (ex ante).
The problem with probabilistic forecasting is that it is almost impossible to work out ex post whether the ex ante probabilities were correct or not. If the forecast is for a 90 percent chance of deflation, and a 10 percent chance of inflation, and the actual outturn is inflation, does that mean the forecast was correct, or was the chance of inflation underestimated?
“RISK MANAGEMENT” DANGERS
In the current circumstances, central banks and investors may be putting far too much weight on the rather distant prospect of an economic collapse.
We have been here before. In the early 2000s, former Fed Chairman Alan Greenspan and then-Governor Ben Bernanke justified keeping monetary policy ultra-loose for an extended period by citing the threat of deflation. That decision is now criticised by some commentators for sparking the subsequent period of excessive risk taking and financial imbalances that led to the banking collapse in 2007-2009.
But Greenspan, Bernanke and many of other rate-setting colleagues continue to defend the “risk management” approach they adopted in 2003-2004, arguing it was correct in the circumstances. The deflation danger was very real, they argue, and successfully averted. In recent speeches, Fed Vice-Chairman Donald Kohn has accepted the Fed made mistakes in bank supervision, but insisted its monetary policy was correct.
Others are less convinced. Kansas City Fed President Thomas Hoenig’s repeated votes on the Federal Open Market Committee to drop the “extended period” language on keeping rates low are grounded in the belief that maintaining low rates for a long period creates distortions and sows the seeds of future crises.
Strategy from the aftermath of the last recession is now being repeated in the United States, the United Kingdom and the eurozone. Policymakers have postponed normalising interest rates after the immediate banking crisis has passed to deal with the lingering “tail risk” of descent into a new recession.
Investors too are transfixed by tail risks. One set is terrified about deflation and defaults, causing a surge out of equities and into the safest U.S. Treasury bonds. The other is worried about a resurgence of price pressures, triggering a flight from cash into hard assets and gold.
Other commodities are wavering between fears of inflation (which is keeping pension funds and other investors interested in being long of commodity futures and hard assets) and fears about recession (resulting in repeated selloffs).
But the point about tail risks is that they lie in the tails of the distribution. Even with fat-tails, tail risks are still fairly unlikely. Having spent much of the last two decades ignoring the fat tails in returns from many assets (“kurtosis”) policymakers and investors now risk over-compensating and focusing on tail risks to the exclusion of the main body of the probability distribution.
When the current bout of panic and gloom passes, it is much more likely the global economy will continue to muddle through, with moderate growth and moderate inflation, than that it will be plunged into a Third Great Depression or some sort of hyperinflationary spiral.
Investors need to take tail risks into account (as some investment banks and hedge funds obviously failed to do in 1997 and again in 2007-2009). But it makes little sense to promote tail risks to the status of “central scenario”.
Sophisticated investors will probably make good money taking the other side of these tail-risk trades opposite their more panicky and euphoric brethren.