More Taper History, And Why the S&P Rallies Overnight

By in
More Taper History, And Why the S&P Rallies Overnight

Two “Data” topics today:

Issue #1: We recently showed you that 2013’s “Taper Tantrum” had no effect on US equities, but what happened the next year? Recall that the Federal Reserve did not actually begin reducing its bond purchase program until 2014. The 2013 “tantrum” – when Treasuries went from 2 percent to 3 pct in just 3 months – was about then-Chair Bernanke’s poor communication on the topic, starting with a Congressional hearing in May 2013.

The following chart of the Fed’s actual balance sheet from 2013 – 2015, indexed to 100 on January 1, 2014, shows this well. Assets grew by 38 percent in 2013, and at a pretty steady clip. Once tapering kicked in, balances grew by just 10 percent in 2014. In 2015, they were unchanged.

Now, here is a 2013 – 2014 chart of what effect the “tantrum” and then actual 2014 tapering of bond purchases had on 10-year Treasury yields (black dotted line, left axis) and the S&P 500 (red line, right axis).

The truth of the matter is that Treasury yields fell from 3.0 percent to 2.2 percent during 2014 and the S&P 500 rallied 11.4 percent for the year. Tapering coincided with good returns for both US stocks and bonds.

Now, the sharp-eyed reader will ask “what happened in October 2014?” – that large S&P 500 drawdown you see on the right side of the chart just above the final date shown, and the coincident drop in Treasury yields. That had nothing to do with tapering. Rather, it was a good old fashioned growth scare, centered on a decelerating European economy. It lasted all of 2 months and hit the S&P 500 for 7 percent before it bounced back to fresh YTD highs.

Takeaway: tapering – actual or merely signaled – did not negatively affect US stocks in 2013 (S&P total return +32 pct) or 2014 (+14 pct). This should not be surprising. The literacy rate among investors is 100 percent; tapering surprised precisely no one. What mattered to equity prices were issues like slowing global growth, which is why 2014 got pretty choppy in early Q4. We believe the same dynamic applies to today – lots of things will matter to equity prices over the next year, but we doubt tapering will be the most important issue in capital markets.

Issue #2: Why do US stocks so often seem to rally overnight rather than during regular trading hours? The New York Fed just today reissued an analysis on this topic, and even though we mentioned it when it first came out in 2020 their data is definitely worth revisiting.

First, here is what the Fed dubs the “Overnight Drift”. The chart below shows average daily S&P 500 futures returns by hour from 1998 – 2019. The time stamps on the X axis are East Coast US time, starting at 6 pm local. As you can see, the largest returns (when the red line goes parabolic) are from 2am to 3am. During regular trading hours (930 to 1600, right side of the graph) average cumulative returns are basically flat.

The Fed attributes this “Drift” to US equity market makers absorbing excess supply during selloffs at discounted prices going into the 4:00 pm close, which in turn creates an artificially low S&P closing price. The chart below shows the “Overnight Drift” broken down by days with high “RSV” (Relative Signed Volume, the percent of sell imbalances) versus those with no RSV or negative RSV (lots of buy interest at the close). Sure enough, days with high levels of selling at the close have a more pronounced positive “Overnight Drift”.

Takeaway: to the Fed’s thinking, the Overnight Drift is caused by non-US traders sniffing out artificially discounted closing S&P prices and arbitraging them away (albeit with some risk) before the next trading day in the States. As for a practical takeaway, the first chart we showed you says that on average you’re often better off selling equity positions at the open and putting in your Buy tickets in the middle of the afternoon. Market events can always subvert that approach, of course, but on average it has been a productive rule set to follow.


Fed Overnight Drift analysis: