Beta release

Ouch. There’s no other word to describe it. Over the last 24 hours, virtually everything that could have gone wrong, has. The dollar bet has gone from super to supine, courtesy of Friday’s inexplicable late-day dollar surge that has thus far followed through on Monday. The DAX, meanwhile, has handily beaten the S&P over the last few days, taking the spread trade (un) comfortably into negative territory.

Macro Man was not particularly surprised that the dollar corrected; after all, December 2004 and 2005 saw corrections from the underlying trend. What was surprising was the manner in which it corrected, with seemingly bullish, Voldemort-driven price action swiftly reversing into a free fall. Ultimately, this could be a healthy phenomenon, shaking out a few weak dollar shorts and setting the stage for acceleration of dollar weakness into year end. Perhaps tomorrow’s Fed meeting could be a catalyst if Jeffrey Lacker returns to the ‘on hold’ fold and the statement is perceived as dovish.

However, the losses prompted Macro Man to put on the thinking cap over the weekend. What is the appropriate ‘default’ setting for a macro portfolio? After all, it is pretty difficult to scratch out 10% in pure prop trades without taking sufficient leverage to risk substantial drawdowns. The recent press accorded to Goldman’s Absolute Return Tracker suggests that it should be possible to create an underlying beta portfolio on top of which to lay a series of portable alpha prop trades.

Of course, there is an argument to be made that that is what many hedge funds currently do, without the portable alpha overlay. Nevertheless, if one can develop a passive strategy that generates a consistent return, it provides a handy tailwind from which to leverage one’s particular area of expertise.

An obvious starting point is equities, which are more volatile than bonds or currencies and deliver higher nominal returns. Macro Man’s interest was piqued by a recent debate about the utility of P/E ratios in market timing. Macro Man decided to investigate a relatively naïve market timing strategy based on S&P 500 trailing PE ratios.

The essence of the strategy is to be fully invested in US equities when they are relatively attractive, and out of the market when they are not. Macro Man used trailing S&P 500 p/e ratios since 1986 (the earliest available in Bloomberg), inverting them to generate a trailing earnings yield reading. He then compared this earnings yield with the prevailing yield on 10 year Treasuries. The initial analysis seemed to suggest that when Treasury yields are substantially higher than earnings yields, equities underperform. When the difference is smaller, equities seem to do OK. What is curious is that Treasury yields are currently priced at a discount to earnings yields, a clear anomaly over the past 20 years!

By this reckoning, equities look cheap. Or, to put it another way, Treasuries look expensive, thanks to central bank and pension fund bids. However, eyeballing a chart is a pretty lousy way to demonstrate a relationship, so Macro Man dug deeper. He broke the UST – SPX yield differential into different buckets, comparing the return characteristics of each bucket. The results, displayed in the table below, appear to demonstrate a clear relationship between the bond-stock yield differential and S&P 500 price performance. (Note that Macro Man compared each day’s price return with the yield differential as of the previous day’s close.)

This is an approach similar to that of the so-called Fed model, except that it does not purport to determine the degree of over/under valuation of the stock market. All this approach does is determine whether the market looks attractive or unattractive. There appears to be a crucial level in the yield differential of around 2%, whereby stocks offer an attractive rate of return below that level and an unattractive rate of return above the level. So Macro Man devised a simple trading strategy wherein he is long the S&P 500 if the yield differential is below 2 and out of the market if it is above 2. The results, displayed below, are impressive. By getting out of the market when valuations look extreme, Macro Man improves the return/risk ratio of the buy and hold strategy from 0.58 to 0.86.

Now, this analysis did not include transaction costs, but then again it did not include dividend returns, either. Macro Man believes that the latter are substantially higher than the former, so if anything the returns on the strategy are understated. Note also that the strategy was out of the market during the 1987 crash and for much of the 200-2202 sell-off. This looks like a fairly reasonable starting point for a macro beta strategy. Over the coming days (and perhaps weeks), Macro Man will look at other passive beta strategies, such as the FX carry trade.