In short, buy stocks.
Remember that 2011 NY Fed study which found that since 1994, a stunning 80% of all equity returns on U.S. stocks were generated over the twenty-four hours preceding scheduled Federal Open Market Committee announcements, a phenomenon called the pre-FOMC announcement “drift.”
Well, we are now in that 24 hour period - which starts roughly now and accelerates into 2pm on Wednesday, when the Fed is widely expected to hike by another 25 bps - and over the weekend Goldman decided to take another look at this performance "drift", or as Goldman calls it, "the FOMC Alpha" and similarly finds that average returns are higher on FOMC announcement days.
So is tomorrow (and today) a buy-and-forget "guaranteed profit" day for daytraders just because "new information" is released from the (overly dovish) Fed? The answer appears to be yes.
As Goldman writes, market participants and policymakers are well aware that asset prices respond to new information released on FOMC announcement days. But, what is perhaps less well known is that a sizable portion of average yearly asset returns is generated on these days. For example, a strategy that goes short the USD versus a basket of all other G10 currencies has provided an average daily annualized return of about 80bp on FOMC announcement dates, compared with an average daily annualized return of about 30bp on any other day of the year.
To make this point, the bank demonstrates that stocks and foreign currencies outperform on FOMC announcement days.
"We show that the ‘FOMC alpha’ has been sizable for long positions in major DM equity markets, as well as for strategies that short the USD versus all G10 currencies but the Yen, and versus some major EMs (TRY, ZAR and MXN in particular). By contrast, investments in major sovereign bonds do not tend to deliver a larger or smaller return on an FOMC day relative to the return earned on any other day of the year. We also find that these patterns are more sizable in the post-Global Financial Crisis (GFC) period."
Still, despite the demonstrable "drift" effect, not even Goldman knows precisely what to attribute it to: "what explains the ‘FOMC alpha’? We need to know this in order to assess whether these observed patterns are likely to persist and, if so, for how long. Among the items listed are compensation for risk, risk appetite, a central bank that surprised more on the dovish than on the hawkish side on FOMC announcement days, and investors’ perception of a ‘dovish Fed’ are non-mutually exclusive explanations for the excess asset returns earned on FOMC days. A few or all of these factors play a role in explaining the excess return on some assets on FOMC announcement days.
What does it mean for the future "the patterns in asset returns observed on FOMC days in the past few years, on average, are unlikely to disappear any time soon" which likely suggests that buying now and all that way until 2:00pm tomorrow is a "risk-free" trade, unless of course everyone jumps on board, in which case it just may not work.
The findings are relatively straightforward: as the first two charts show, while equity markets tend to outperform across the board on FOMC days, the USD is to be shorted.
As Goldman writes discussing the above charts, which show sizable excess returns on FOMC announcement days, "equities and foreign currencies have tended to generate positive excess returns on FOMC days (Exhibits 3-4)."
In particular, on FOMC days, the S&P 500 has historically moved up by an average of 30bp and other G10 equity markets indices have also delivered positive returns. These returns are statistically different from average returns on any other day of the year. Turning to currencies, the USD has tended to depreciate versus all other G10 currencies, except for JPY. In EM FX space, we concentrate on the most liquid currencies (BRL, MXN, TRY, RUB and ZAR, see Global Markets Daily: EM FX sensitivities ahead of DM central bank meetings, Sep. 9, 2016 for more detail) and find that they have all outperformed the USD on days when the FOMC meets (Exhibit 5). In general, across G10 and EM FX, currencies with a wider interest rate differential versus the USD have provided higher returns (Exhibit 6). This suggests that a portfolio that goes long high-yielding currencies and short the USD could potentially deliver the best relative performance on an FOMC day."
The next set of charts show the recurring beating which the dollar tends to suffer on FOMC days:
It goes without saying, that excess returns on stocks and foreign currencies tend to be larger (smaller) on a FOMC day when the Fed delivers a dovish (hawkish) surprise on the day.
And while the drift has been demonstrated to exist since at least 1994, the "FOMC Alpha" has been larger post crisis. Goldman explains:
Given the sizable average excess returns on some investments on FOMC announcement days, it is important to understand whether the patterns we have identified so far are robust, or are driven by only a few FOMC meetings. First, we split the sample into two sub-periods: pre- and post-2008. We note that the patterns we have shown above are more pronounced since 2008 than before (Exhibit 7 and 8). Second, we check to see whether the excess returns on long equity and short USD positions on FOMC announcement days hold even in a more recent sample, starting in 2014. They do. However, given the limited number of FOMC meetings included in this shorter sample, the difference in returns on FOMC and non-FOMC announcement days is not often statistically significant.
Finally, we compute rolling average excess returns over a one-year period (and, hence, only eight FOMC meetings) to check that no particular period drives the difference we find between returns on FOMC and non-FOMC dates. Exhibits 9 and 10 show the rolling coefficients for the S&P 500 and the trade-weighted USD index we compute versus all other G10 currencies. Even though the period 2008-2009 drives a lot of the action, the average change in FX and equity prices on FOMC announcement days is larger than that for other days across the entire sample.
One striking observation from this exercise is also how much of an outlier the 2013-2014 period was. Starting with the June 2013 press conference in which then-Chairman Bernanke laid out a ‘taper timeline’, the FOMC delivered an uncharacteristic series of hawkish surprises. That included in March 2014 an upward revision to the ‘dots’ and remarks from Chair Yellen that the first hike could come just six months after the end of QE, as well as a number of smaller hawkish surprises.
Our takeaway from this analysis is that this period was quite unusual for the Fed, and probably partly the result of a series of communication missteps rather than a genuine effort to ‘be hawkish’. Through this lens, it is understandable why the Fed is now trying so hard to communicate more clearly around the start of balance sheet normalization. It also demonstrates that, even now during this period of policy tightening, market participants who expect the FOMC to deliver hawkish surprises to tighten financial conditions may be disappointed, at least on announcement days, as that is not typical.
But the best visualization of the "FOMC Alpha" is laid out as follows: To create the underlying data, Goldman focused on meetings with a press conference starting from 2013, particularly when the Fed releases the Summary of Economic Projections and the Chair holds a press conference.
The chart below show the average change in the EUR/USD and the S&P 500 cumulated minute-by-minute, starting at 8.30 am NY in the morning of the FOMC day until the NY closing. The S&P 500 continues to move higher during the press conference, a testament that on average press conferences continue to deliver ‘dovish’ news. Movements in the EUR/USD have been less clear-cut, on average, between 2 pm and the end of the press conference, but the evidence after the press conference that the USD continues to depreciate on average also points in the same direction.
The obvious conclusion: a sizable portion of excess returns on riskier assets is generated, on average, on FOMC announcement days rather than on other days of the year. In addition to the academic literature, which has identified this pattern, Goldman has likewise shown that it still holds in very recent periods. That said, the bank makes it clear its analysis is not about one single FOMC meeting, when the Fed was particularly hawkish or dovish, nor it is about ‘getting it right’ on any single meeting but instead to highlight the oddly recurring "investment opportunities that are more likely than not to be profitable over a series of FOMC meetings during one or a number of years."
So will buying the S&P now and all the way until 2pm tomorrow be another profitable trade, or - now that everyone has been reminded of this "risk free" trade - will this self-referential perspective spoil the predestined prfotable outcome? The answer will be revealed tomorrow shortly after 2pm.