August post-mortem

Well, that was an interesting month! To say that a bevy of (black) swans descended on financial markets would seem to be something of an understatement, what with the implosion of the money markets and the explosion of some quantitative investment strategies, most notably in equities.

And yet...the simplest of strategies, buy and hold a 60/40 mix of equities and government bonds, would have yielded positive returns in both the US and Eurozone. That the S&P 500 and DAX both finished higher in a month when markets drastically shifted their monetary policy expectations towards easier conditions as a result of financial distress is a testament to how bizarre August really was.

August was a good month for Macro Man- his best of the year in both the blog portfolio and, more importantly, the real world. His 2.88% return, while probably not comparable to those positioned aggressively short certain segments of the credit market, nevertheless was quite a pleasing development. To have made money last month, on top of solid profitability earlier in the year, demonstrates the importance of remaining flexible in one's approach. While Macro Man has many flaws, he is fortunate that a reluctance to change his mind is not one of them.

In line with the peformance of the SPX and DAX, Macro Man's beta portfolio registered a modest positive on the month. While the equity beta plus strategy is technically a quant model, it is an exceedingly simple one. In developing quant models, portfolio managers have two options: make it as sexy as possible, and run the risk of data mining (and incur the associated 26 standard deviation negative event), or pursue a KISS strategy: keep it simple, stupid.

While KISS leaves a lot of potential information on the table, in Macro Man's view it is also less likely to end up with a convoluted portfolio and a false sense of security. There is of course no guarantee that a KISS strategy will not incur a disastrous left-tailed return; then again, Macro Man knows that and acts accordingly in his alpha portfolio. And in a month of noise and sexy-model dislocation, the KISS beta plus portfolio returned 0.80%.
Strangely, the equity alpha portfolio lost money on the month. While there was a fortuitously-timed sale of SPY and DAX option expiry, the return on the Brazilian equity position was sharply negative. This was slightly galling, because the original trade construction proved to be correct (hedging EWZ with $/BRL calls), but Macro Man neglected to roll his FX option position, leaving him with a naked long Brazilian equity exposure. This was a costly price to pay for lethargy. All in, the equity alpha portfolio dropped 0.23% on the month, leaving the overall equity P/L positive on the month.
Fixed income alpha trades dropped 0.10% in August. The primary risk position was long TIPS hedged with JGBs rather than UST, and the outperformance of JGBs on the month proved costly. However the JGB losses were partially offset by the uptick in SGD swap rates, so overall losses were modest. The SGD swap is approaching the initial target levels that Macro Man had on when he first did the trade, so he may look to take profits some time this month.
Finally, the FX alpha portfolio delivered a solid 2.42% positive return. This was largely a function of a large long implied volatility position, and validated Macro Man's views that a) asset market volatility was likely to be correlated, and b) that FX vol represented an excellent cheap hedge. While profitability of the long vol position is well off its intramonth high, and could well sustain a further drawdown over the next week or so, Macro Man remains of the view that implied volatility across assets is now a buy on dips rather than a sale on rallies (which has been the case for the last 4 years.) While this does not preclude Macro Man from taking profits on strength, he has little inclination to sell volatility after most of the move has been retraced.
All in, the alpha portfolio returned 2.08% in August. Finally, all the worrying paid off! August was especially satisfying given the relatively dreary performance of many other asset managers- the chart below shows the returns of some pure FX traders (white) and macro managers (orange.)
The swift collapse in risk-asset strategy returns since mid-July should convince you of two things:
* Risk premia are not, in fact, dead
* The best time to hedge is when it looks as if you least need it
Previous
Next Post »

6 comments

Click here for comments
Anonymous
admin
September 3, 2007 at 12:27 PM ×

Good job, MM.

Perhaps you could ask your boss for monthly holidays then? Comparing your performance in August with mine the only thing that keeps me from using you as an example with my boss is the fear of him liking the idea so much he decides to give me permanent holidays... :)

avinash goldfish

Reply
avatar
Macro Man
admin
September 3, 2007 at 12:49 PM ×

I can still be of use to you, though, Avinash. You can use the execrable returns on my Brazilian RV spread as either a favourable comparison (see, I did better than this clown) or an excuse (this guy did OK everwhere but Brazil!)

Sadly, when I proposed the "2 weeks of holiday per month" strategy to my boss, he thought I was joking....

Reply
avatar
Anonymous
admin
September 3, 2007 at 4:15 PM ×

Well done MM.

I compliment you on your transparency.

I for one run a hedge fund but more recently have moved away from monthly performance feedback.

No I am not losing money in fact I am doing pretty well, however, my style of trading (contrarian) just does not produce returns in a linear month-by-month fashion.

I have recently taken to reading your site daily, and I am really impressed with the quality of your writing.

Reply
avatar
Macro Man
admin
September 3, 2007 at 4:33 PM ×

Thank you very much for your kind words. I also believe that high frequency performance reporting is usually countrproductive and encourages suboptimal investment decisions. You may be interested in this piece on the subject that I wrote a few months ago.

I try to publish a daily P/L for a couple of reasons: I think people are generally interested in quasi-real time performance information (and the concomitant emotional angst that it generates), but more importantly (for me at least) is that I use the portfolio as a laboratory to try things out. Keeping a daily record helps me to assess the efficacy of current or potential strategies for my 'real job'.

Reply
avatar
September 3, 2007 at 6:17 PM ×

Right on about KISS and mechanical systems. Find something that tells you when to be in, in general terms, and something else that is intended to keep you out of periods of volatility during that time - and if the backtest gives lots of cushion and isn't too finicky about optimal lookbacks and the like... Then, if it's still workable over a couple of days of lag in execution, you've probably got something interesting.

Reply
avatar
Victor
admin
September 4, 2007 at 3:44 AM ×

Well done, achieving 2.88% in a month full of swings and volatility sure isn't easy. Keep it up!

Reply
avatar