Quote from Ghost of Cutten:
Nice debate on broad vs specific exposure. Allow me to provide a more specific example of the benefits of simplicity vs complex/targeted exposure. A good illustrative example was commodities and financials during the 2007-09 bear market. Financials rolled over a lot earlier than the general market, and were down huge in 2007, whereas commodities kept going great guns up to H2 2008. Some made very nice returns in 2007 to mid 2008 by a simple long commodities short financials spread, whilst the S&P was not down that much. For example, from the Oct 2007 peak, XLF fell 1/3 up to early July 2008, whereas the RJI commodity ETF rallied 40% over the same period - a 73% return on the spread trade. The spread then lost 43% in the next 7 trading sessions, mainly due to a huge 35% bear market rally in the XLF.
In individual stock names it was even more extreme. Fannie Mae went up 170% in those 7 days, despite being ultimately destined for $1 per share and below.
By contrast, SPY rallied 7.6% during that period.
So, let's compare the total return to max drawdown ratio of all these plays:
i) simple bear play - short SPY from peak (Oct 2007) to trough (March 2009). Total return: approx 55%. Max DD: 27%. Ratio: 2:1
ii) sectoral bear play - short XLF from peak to trough. Total return: 85%. Max DD: 60%. Ratio: 1.42:1
iii) stock-specific bear play - short FNM and LEH from peak to trough (I am being favourable to stockpicking here by choosing 2 ideal shorts that went to zero or near enough to it). Total return: 99%. Max DD: 120% (this DD took place in 3 trading sessions, lol). Ratio: 0.825:1
iv) long/short thematic play - I already showed the results from short XLF long RJI above. Unless you somehow ran the trade for 1 year then exited near the peak, you were looking at making about 1:1 return to max drawdown here, maybe up to 1.5 if you were nimble.
So, let's compare the stock-picking thesis against this data. The two most perfectly selected shorts of the greatest crisis of the modern era; the best long/short 'market neutral' sectoral macro spread play during that time; a more basic targeted sector short in financials; and a simple plain vanilla short SPY, which would have taken all of 2 seconds to analyse. Note that these are all picked with 20/20 hindsight, which gives a huge artificial benefit to the stockpicker thesis - in reality there's a very real risk that your stock or sector picks actually suck or just perform in line with the sector or market, yet with far more risk. So, these results make the stockpicking approach look better than it really is.
Now, if the 'stockpicking is optimal, simplistic bets are lazy' theory is correct, what would you expect? The stock shorts, after extensive research, should have provided the best return to risk ratio. The spread trade should have provided the 2nd best R/R ratio, as it was hedged against most grey swans and caught the weakest and the strongest sectors in the market over that period. The targeted XLF short should have been the 3rd best, and the boring, lazy, simple SPY short should have provided the worst returns relative to risk.
What actually happened? The complete opposite! Boring, lazy short SPY had a fantastic 2:1 R/R ratio, not to mention the least exposure to grey/black swans.
Relatively simple short XLF made 1.55 times more than short SPY but at the cost of a monster 60% drawdown over twice as big as that in the overall market. SPY was superior on a R/R basis by a factor of over 1.4.
The spread play turned out to be a Texas hedge, dropping 42% in a week and a half while the market only budged 7.5% i.e. it was *at least* 6 times riskier than simple short SPY, yet made only 1.3 times as much - in other words, it was a shittier trade by a factor of 4.5 times.
And the perfect short - two individual stocks falling 99-100% each and ending on the pink sheets within a year - well, quite apart from the buttock-clenching 120% 3 day face-melting sucker rally (and another near 100% spike in Spring 2008, again taking only a few trading sessions), if you somehow managed to hold onto your shorts, then you made less than your risk. You dropped 120% to make a 0.85 R/R ratio, comapred to dropping 27% in SPY to make a 2:1 R/R ratio. This makes the stock-specific ideal shorts a worse R/R trade by a factor of 10.5.
So...plain SPY short was the best trade, and over ten times better than the ideal 2 stock short play. The simplest trade had the best R/R ratio and the lowest outright risk. Each additional layer of complexity made things worse.
Clearly, the only way to really earn superior returns by stockpicking here was to buy puts on the short plays (or buy CDS as Einhorn did). But, did Einhorn make 55% on capital with his Lehman play? No he didn't - in fact he was down substantially in 2008. He got the stock pick right but got hosed because he got the big picture call totally wrong.
Now, as darkhorse rightly points out, occasionally you get a flat market where one or two sectors or stocks are on fire. But you also get flat markets where the same sectors or stocks get hammered. Is anyone really nimble enough to switch between the two? Maybe. But on this occasion we were talking about an obvious macro play (shorting an index), versus a complex riskier one (shorting 2 stocks with more specific exposure). The 2008 example shows that the latter might not just take more time and effort, but is quite likely to provide more outright risk and a significantly inferior R/R ratio. That is why simple exposure should never be dismissed as the inferior, lazy trader's option. There is much virtue in simplicity.
In my opinion, stockpicking is best for when there is little or no macro opportunity. You can't earn alpha by big picture calls in a quiet market, and by having a long/short trading book you are earning alpha on both sides whilst massively reducing risk (and you can still dial in the index exposure of your choice via futures). But when you have a slam dunk macro call, the argument for stockpicking is weaker than you would expect.
I would say the best use of stockpicking in big macro plays is for finding the best asymmetric payoffs via options, CDS, and (to some extent) spread plays for their lower overall market risk (although you have to watch out for when these become crowded and start turning into Texas Hedges). Those make more sense to me for targeted exposure than just shorting the stocks themselves. Another lesson from 2008 is that getting the macro right and then finding & picking the best trading vehicle (financial/RE CDS positions) earns a heck of a lot more than just getting the macro right (short SPY).
Summary - simple broad exposure is a very robust, low risk, and surprisingly attractive play. Doing a bit more legwork to pick specific sectors or stocks might actually backfire due to the huge increase in risk. Probably the best play is to pick not just the stocks but the best trading vehicles for them. I personally would prefer the bulk of my exposure in a broad index or (if its just a sector that's going to move) sector ETF, and then a few small speculative asymmetric bets on targeted individual stocks which best express the macro view.
). Conviction is somewhat tenuous even on the best trades. 