On 4 January, Federal Reserve chairman Jerome Powell stated that

“The markets are pricing in downside risks … and they are obviously well ahead of economic data”.

Admittedly, the signals from the financial markets are not very encouraging. Since the end of September, Wall Street has dropped sharply and the spread between 2-year and 3-month rates has narrowed dramatically. In the past, this situation has generally been followed by recession, albeit with a lag that has varied in length. “In general” means that some of these signals were wrong, and Mr. Powell’s statement suggests that this might now be the case.

In the light of recent economic data, the markets seem to be extremely pessimistic: the nowcasts of the Atlanta, New York and Saint Louis Federal Reserve banks are calling for annualised Q4 2018 growth of 2.8%, 2.5% and 2.6%, respectively (these figures factor in December’s strong job creations and the drop in the ISM indicator).The Federal Open Market Committee members are projecting a 2019 growth of 2.3%. One explanation for this dichotomy between the real and financial spheres is that investors and economists are making different economic analyses, even though both are using the same economic data.

Another explanation for the decline in the equity markets and long-term rates at a time when economic statistics are robust is that different horizons are being used: the markets tend to anticipate trends several months in advance. Yet if this is the case, and growth is currently as strong as it seems to be, then which factors are likely to trigger a sudden slowdown over the next few months? A slowdown is even harder to imagine given the inertia of growth figures, which tend to evolve slowly, especially at a time when monetary policy is very cautious, reducing the risk of an abrupt interest rate shock.

A third explanation is that while economic forecasts tend to favour the most probable scenario or the weighted average of several possible scenarios, the markets are more likely to focus on extreme risk. The level of long-term interest rates does not depend solely on real growth and inflation expectations, but also on divergences in these factors, possible asymmetries in their respective distributions, abnormally high extreme risks, and even the correlation between the bond and equity markets. The perception that economic news is more likely to be bad than good (the US expansion phase has reached a respectable age), the presence of major extreme risk (US-China trade war) and the negative correlation between bond prices and the equity market (a classic phenomenon for several years) will drive down long-term rates and contribute to the flattening of the yield curve. This is an alarming tendency as revealed by the Fed’s recent survey of senior loan officers. They effectively claimed that an inversion of the yield curve would push them to tighten lending criteria because historically, inversions have been followed by recessions. Past experience is thus influencing today’s behaviour, and feeding self-fulfilling fears.

A fourth explanation for the dichotomy between the real and financial spheres is the complex relationship between principals and their agents. The first, the holders of financial assets, may be institutional investors, private banking customers, etc. The second comprises the institutional investors’ management teams, asset management firms and mutual funds. Traditionally, both types of players display risk aversion: asset holders are more sensitive to losses than to the opportunity cost linked to insufficient exposure to a bullish market, while their agents do not want to risk losing their jobs, or their assets under management, in case of a poor performance, whether in absolute terms or relative to their peers. As a result, risk aversion is squared: when uncertainty rises, they tend to react disproportionately to economic news. Clearly, it would be dangerous to adopt a very narrow interpretation of market data that looks solely at economic signals. Yet we cannot ignore these signals either, as biased as they may be. As so often is the case, it is all a matter of nuance.