Aug 3, 2016

Update on a Systematic Alt Risk Strategy

Early this year I wrote a post on self-managed Systematic Alternative Risk strategies for retail investors.  I had been managing an amateur version of a strategy in that area for a while and had pitched some its benefits for the right kind of investor.  The general idea at the time was that if one were to have the background, inclination, and a rudimentary set of skills necessary to manage systematic rules-driven risk/return, I didn't think it was necessarily a bad bet going into a what sounds like the cloud of crummy expectations about future returns that is in the air these days.  Here is an update on the strategy and some comments on where I think I'm headed now.



The chart below is the current mean-variance view of my self-run systematic alternative risk strategy. The strategy is mostly a rules-based system focused on fixed-income momentum. In addition, it tends to have an expansive view on how collateral yield is managed plus it makes a few discretionary macro and short vol trades here and there. The time frame chosen (Jan 2014 - July 2016) is probably somewhere between arbitrary and cherry-picked and is modestly unfair to the other-fund comparisons I am trying to make. This is also pre-tax by the way; after-tax would likely not be very flattering to me and who likes to un-flatter themselves in public.

The green dot is me. The white dot is an asset allocation tracking ETF that attempts to follow the 'S&P Target Risk Moderate Index.' This ETF, for me, is supposed to be a proxy for a core, conservative asset allocation portfolio, something that the index/ETF happens to not have tracked very well in my opinion. The red dot is a large, popular momentum-focused liquid-alt mutual fund. It is used for comparison since I tend to be heavy on momentum and systematic rules-based trading. It comes from a firm (this is in code but easy to decode) that is leading attempts to "translate academic insights into quantitative strategies for institutional and retail investors," i.e., this is a current Goliath in the systematic liquid alt-risk world. This is a firm against whom I should probably not compare or contrast myself. They are a lot smarter than me, I have a lot of admiration for what they do and how they do it, and I also invest in a few of their funds.




I've been running this thing since about 2011.  Do I have any new conclusions since my last post?


  • I'm still more or less on track with my goals on a pre-tax basis (up and to the left)...but on this time frame only; I did mention cherry-picking the time frame, did I not?  I would prefer to have at least 40 years of data but I have what I have. A longer term view would likely bury me in under-performance on this map.  I don't think that in any honest world I will be able to outperform the market over the long run and I am not expecting to do so, except over selected very short time frames and under very unusual circumstances. 
  • My personal sense is that I would now really prefer, some of the comments above and below notwithstanding, to have a long term goal to move everything towards passive indexing: it's cheaper, easier, more tax efficient, and can achieve similar or better results...unless I can unlock some alchemic break-thru sometime soon.  
  • Tactically speaking, on the other hand, I feel like there might be some distinct advantages to this kind of strategy in the low return, high volatility environment that everyone seems to be predicting for the next X years.  So, for now I'll hold off on shutting this down in order to see what happens. 
  • In addition, even though I didn't post the correlation matrix, this strategy still has -- with respect to the major traditional asset classes -- some OK diversification (non-correlation) benefits as a small component within a larger core portfolio. It also  has a a pretty good track record of low-vol positive returns. These return and correlation expectations are something I am comfortable projecting into the (near) future. On the other hand, I realize that some of this positive-risk-premia-low-vol-low-correlation stuff is achievable, to a certain extent, elsewhere, maybe. The mindset, however, will definitely inform my allocation decisions going forward, whichever way I go.   
  • I haven't looked at this for a while and it is not shown here but the strategy returns were at one point decently asymmetric: there was less downside than up. That's a good thing for me and something I'd like to replicate even if I ditch this strategy in favor of a fund manager. The asymmetry in this case was measured with what's called the "Omega ratio." It calculates the ratio of the "upside area" under a cumulative distribution function of periodic returns divided by the "downside area" under the cumulative distribution where up and down are split by some arbitrarily selected threshold level.  The threshold is the trick, of course, but mostly I think the results looked pretty good the last time I ran the numbers.  I'd hate to give up that asymmetry...if I can keep it going. The strategy/ratio, for what it's worth, has never been significantly tested by a major correction or an extremely adverse event in rates. 
  • At the very least the strategy is completely liquid and mostly free (last year was 3-5 bps after being closer to 20 the year before). There are no lockups, no gates, no LP agreements, no K1s, no opaque counter-parties, and no 2 and 20.  Plus I get to manage my own downside risk. I like that. It is hard to overstate this sense of freedom from 2nd-or-3rd-rate private-placement fund managers (I don't have access to the "1st rate" managers for whom I might actually be willing to reconsider some of the "nos" above). 
  • In the end, all of this -- the strategy, the trading, the analysis, the risk, the time and effort, the tax reporting, and sometimes the fear --  seems like an awful lot of work for some fairly subtle benefits. It doesn't feel like a lifetime strategy to me at this point. On most days it feels more like riding a bike in first gear: a whole lot of motion for very modest forward progress.  It's too bad I can't have someone else run this thing for me...but only under the terms of the previous bullet point.




No comments:

Post a Comment