CTAs vs Central Banks: The QE Effect

September 7, 2018 01:32 PM
The worst performance periods for CTAs corresponded with the various quantitative easing regimes of the Fed. Is this a coincidence?
CTAs vs Central Banks: The QE Effect

CTAs vs Central Banks: The QE Effect

In January 2015 at a Managed Funds Association conference I presented findings on the apparent relationship between commodity trading advisor (CTA) drawdowns and the Federal Reserve balance sheet, specifically the outright held securities part of it). The theme of the panel, and a subsequent research paper was: “Managed Futures and CTAs — Where are We and What's Next?”

The CTA industry was undergoing one of the worst and most prolonged drawdowns in its history at that time, and many people were looking for a feasible explanation. Some have suggested that quantitative easing in all its phases: QE1, QE2, operation twist and QE3 — which was in the ending of phase — could have been the reason due to distortion of asset prices it caused. There was a lot of anecdotal evidence of this, but this possible relationship has not been studied in depth.

Using HFRI Macro Systematic Index as a proxy for CTAs and S&P 500 for equities, we plotted a few charts and ran some numbers, which revealed very interesting findings:

  • There is strong correlation between depths of the drawdowns and Fed balance sheet amount;
  • Equities benefit in months of active quantitative easing (QE) and suffer otherwise; and
  • CTAs suffer when QE is on and do well when QE is off.

In other words, during the QE 1, 2 and 3 periods, CTAs lost money on average, and equities made money hand over fist. Between the periods the picture reversed, equities were losing and CTAs winning (see chart below). The effect was so strong, that if measured via so called statistical hypothesis testing approach, then the neutral hypothesis of no relationship between Fed activity and CTA performance would have to be rejected at 5% significance level, a very strong result for a finance-related analysis, and at 1% for same relationship with equities – an even stronger result, which is quite unusual in financial data.

The presentation and the article have generated considerable interest. And if you accept the findings that QE had a direct effect on CTA and equity performance, it is worth following up now that the Fed is preparing to unwind its huge balance sheet. 

So we ran the numbers again, with a few small changes. First, I have used AlphaBot for performance and drawdown analysis. AlphaBot is designed for a quick evaluation of performance across different providers (funds, platforms, funds of funds, databases, etc.) and asset classes, from hedge funds and CTAs to equities and even cryptocurrencies. Second, instead of HFRI Macro Systematic Index we use the BarclayHedge CTA index. The indexes are very close, but BarclayHedge makes their indexes available on AlphaBot free of charge.

After running the CTA index through the drawdown analysis tool we notice that the CTA industry continues to struggle. Almost as soon as the previous drawdown ended in January 2015, the longest drawdown in CTA index history, a new one began three months later (see table below).

These two drawdowns account for almost seven years of under-water performance for the industry, and their length is unprecedented, both individually and combined.  

Before we scare all the existing and potential investors away from the CTA space, it should be noted that despite the severity of the CTA drawdown as viewed by the alternative industry standards, it was just a fraction of the largest drawdowns in equities — roughly one third in depth and half in length — even after adjusting CTAs to match equity volatility. Using AlphaBot it is easy to put the S&P 500 and CTA indexes together to produce a telling drawdown chart (see chart of drawdown comparison,below). To make a fair comparison, we matched CTA volatility to that of equities (14.2% annualized for the period), using AlphaBot’s easy to use Volatility Adjustment feature.

So what exactly is happening in the CTA space? It would be difficult to believe that with so many talented portfolio managers, programmers and traders, they have massively forgotten their skills. It would be much more feasible to look for an external factor (or factors) that may influence the performance in the space and this is exactly what we are attempting to do.

Right about the second half of 2014 the Fed has begun implementation of the tapering program, and the outright held securities balance sheet growth has begun to slow down. Miraculously, for a brief few months around the same time, CTAs sprang back to life, generating 12% in the 12 months from April 2014 to March 2015. This was the second strongest rolling 12-month result for CTAs in more than 10 years, bested only by the 17.5% return from April 2007 to March 2008 (see CTA 12-month performance charr below).

But the joy did not last long, and in a few short months CTAs in general have hit another prolonged drawdown. With the Fed out of the picture, many have wondered what is happening and why? The potential cause is not difficult to spot, at least for those of us who accepted the idea that Central Bank activity can distort markets. This time it was the European Central Bank, not the U.S. Fed causing problems.

After some digging around, we have found some interesting bits of information in the ECB balance sheet. The below chart shows growth in the securities of euro area residents denominated in euros. Notice how it exploded just as the Fed began to taper. The securities of euro area residents denominated in euros has grown more than 10-fold growth from 2013 to 2017.

Using this new data we have updated the original analysis, adding the ECB’s metric “securities held for monetary policy purposes” (see chart below). Notice how the ECB balance sheet expanded dramatically just as the Fed’s was balancing off.

So just when the Fed was tapering, ECB has picked up, and the effect of deepening CTA drawdown along with growing balance sheet appears remarkably consistent. To update our original hypotheses testing stats, we have combined the FED and ECB balances (securities portions of each) and ran the numbers. The CTA Index had an average -0.05 returns in months QE was on and 0.6% return when QE was off. The S&P 500 produced returns of 1.21% when QE was on and -0.8% when QE was off. The correlation of CTA drawdowns and the Fed balance sheet was -0.6; the correlation of equity drawdowns and the Fed balance sheet was 0.69.

Simple Hypothesis Testing

                                Barclay CTA Index            S&P 500

                n1                           85                           85

                n2                           39                           39

                Z                              1.12                        -2.21

                P                             0.13                        0.01       


Please note that we had to interpolate the annual data from the table into monthly, and also use monthly euro/U.S. dollar (EUR/USD) currency pair rates to convert the euro-denominated balances into dollars to perform correlation analysis. This creates some smoothing of the data, and may have some influence on the statistics. Regardless of the shortcuts we had to take, the relationship remains remarkably strong, and p-values were low enough to seriously consider the possibility of a casual, and not just statistical relationship.

Some say that this effect may be an example of a “spurious correlation” – meaning it is all just a coincidence or that there may be some other factor influencing the time series. There are four reasons that is unlikely:

  1. We are looking at variables in the same physical realm; finance, and more specifically, asset prices. These are gears within the same gear box, and when one moves it is natural to expect the other one move too.
  2. The timing of upside and downside moves is impeccable. It does not just fit some general timeline, it fits almost perfectly.
  3. The effect is confirmed by two different asset classes: Managed futures and equities, and their relative moves are very telling.
  4. The Central Banks, and not the CTAs, are calling the shots on intervention and amounts of asset purchases. When that money propagates through the portfolio chain, we see prices move in almost all asset classes.

It is reasonable to assume that an influx of trillions of dollars will affect asset prices across asset classes? That was the point of the intervention to begin with? That has been acknowledged by the ECB officials. ECB Executive Board Member Peter Priet noted when describing the benefits of the ECBs asset purchase program that in addition to depressing interest rates, “Higher asset prices spill over, through portfolio rebalancing, into non-targeted markets as investors move up the risk and maturity ladder.”

That phrase is a perfect one-sentence summary of our research. So what does it all mean?  And can this information be used to prepare for what may come next?

Well, there is good and bad news. The good news is that we can hope that when Central Banks have done enough intervention things will return to normal (in general and in the CTA world, in particular) and traders and their investors will be able to generate some impressive performance. It is a big question on when exactly that will be, and what state the economy is going to be in when it happens. Current developments feel more like a spinning out of control addiction, and the more this behavior continues, the more painful it will be to get off it.

The bad news is Central Bank intervention on the scale observed in recent years has never happened before. We simply do not know how things would have played out without the intervention; just like do not know when it will end and what will happen when it does. Another market crash? A spike in volatility? An unrelated event that may bring down the global economy? A simple return to the old normal?

All of these possibilities are likely in some degree, but one thing is certain: the Central Bank activity does affect markets, and the influence is not always positive. It may be better overall if the intervention activity is reduced, and, eventually, stopped, to help everybody catch their breath and return the sense of normalcy and responsibility to the markets.

Dmitri Alexeev, Ph.D. is CEO of AlphaTech Investment Solutions, LLC, creator of AlphaBot.

About the Author

Dmitri Alexeev, Ph.D. is CEO of AlphaTech Investment Solutions, LLC, creator of AlphaBot.