This is because there is a remarkable pattern in stock market returns, uncovered by Anna Cieslak of Duke University and Adair Morse and Annette Vissing-Jorgensen of the University of California at Berkeley. They show that US shares do far better in even-numbered weeks after an FOMC meeting than in odd-numbered weeks. In fact, they say, since 1994 all of equities' outperformance of cash has come in those even weeks.
For example, in week zero (defined as running from the day before the FOMC meeting to three days after it) the MSCI US index has outperformed one-month Treasury bills by an average of 0.57 percentage points since 1994. The following week, it has underperformed by an average of 0.17 percentage points. The week after that, it has outperformed. And so on.
| How the Fed affects US equities | ||
|---|---|---|
| Week | Days since FOMC meeting | Excess return (%) |
| 0 | -1 to 3 | 0.57 |
| 1 | 4 to 8 | -0.17 |
| 2 | 9 to 13 | 0.30 |
| 3 | 14 to 18 | -0.17 |
| 4 | 19 to 23 | 0.42 |
| 5 | 24 to 28 | -0.12 |
| Excess returns are MSCI US index minus one-month Treasury bill rate. Based on weekly data since 1994 | ||
Because international markets are highly correlated with the US, the same pattern exists within non-US developed markets and in emerging markets.
If this pattern continues then equities are due for a bad week. It is now (Friday 27 November) 24 trading days since the last FOMC meeting.
There's a reason for this pattern. It's all to do with the arrival of news about interest rates.
We can think of the relationship between stock markets and news as being one of tension and release. Before news comes out, markets get tense and nervous so prices fall. After the news comes out, this tension is released, so prices rise. To put this another way, prices must fall before news to reflect a heightened risk premium. When the news comes out, investors earn this risk premium. Of course, the news will sometimes be bad and so prices will fall further. But, on average, surprising news is as likely to be good as bad so the surprises net out, leaving us with the fluctuating risk premium.
It's easy to see where the news comes from in week zero, when the Fed announces its policy decision. But why do we get the alternating pattern in the following weeks?
It's because FOMC meetings are not the only meetings about monetary policy. Roughly every other week the Fed's board of governors meets to discuss discount rate requests from the regional Feds. These meetings discuss monetary policy generally. And the flavour of those discussions is released informally to favoured journalists and other contacts. These releases are more likely to come in even weeks - generating the pattern we see.
The Fed justifies this by claiming that it does not want to surprise markets and cause unnecessary volatility, but its informal briefings run the risk of encouraging insider dealing.
Professor Vissing-Jorgensen and colleagues estimate that if an investor could trade costlessly on this pattern - say by moving in and out of an exchange traded fund every other week - he would earn risk-adjusted returns of twice those of a buy-and-hold strategy. This is comparable to those made by Warren Buffett.
We ordinary investors can't do this. But nevertheless, the pattern matters. For one thing, it implies that if you are changing your equity weightings for other reasons, you should try to buy at the end of odd weeks and sell at the end of even ones. It also means you shouldn't read much into modest rises in global markets in even weeks or falls in odd ones, as these are merely the normal pattern.
All this raises a big puzzle: why does monetary policy matter so much? It could be because markets are excessively sensitive in the short term to news and rumours. But there is another possibility, discussed in a new paper by Helene Rey of the London Business School and Silvia Miranda-Agrippino of the Bank of England. "US monetary policy is a major influence on credit conditions worldwide," they say. They show that unexpected rises in the Fed funds rate lead to "significant decreases" in the leverage of global banks. And this is turn leads to lower GDP, rises in credit spreads and lower share prices.
Of course, actual moves in the funds rate are rarely surprises: certainly, a rise next month wouldn't be. But they are only unsurprising because the Fed signals its intentions so much. And it does so in the bi-weekly pattern described by professor Vissing-Jorgensen and colleagues. And perhaps markets respond as strongly as they do because they fear that changes in interest rates will eventually have big effects.
