Fighting Algos

I’m back!  Sometimes it’s nice to take a little time off from the markets to get some perspective.  Unfortunately, I picked a multi-week stretch where the markets started to go nuts with Trump trades.  I suspect there may have also been stops in some trades. 

I got a chance to talk to many traders while I was away and the overwhelming theme seemed to be that “trading gets more difficult every year.”  While that is to be expected in the negative sum game that is STIR futures trading, the pace of technology may be changing the environment at a faster rate than in the past.  I’ll give you some examples of observations I’ve had about how the markets trade a little differently in just the past year:

1.      There is little movement in the yield curve past the greens.  Whatever volatility there was in the long end (with respect to flies moving intraday) has noticeably reduced.  Trades that were profitable earlier in the year require a much longer holding period for smaller gains.  I tend to think that with everyone hiring quants and programmers, the time would eventually come where any kind of curve noise smoothing P&L would mostly disappear.  So that can’t be a longer-term edge.

2.      A curve move that may not make sense can hold for a really long time.  If a few algos are programmed the same way, a kink on some part of the curve could last longer than you would think and/or a kink in the curve could get way worse than you would think.

3.      There is very little volatility in how the curve moves intraday.  It’s somewhat less likely to get a major change in the shape of the curve move intraday.  For example, once we start bull steepening led by SFRM5, it more or less just stays that way all day (even on some days where some secondary data may indicate otherwise).  I’m not saying there aren’t days where the curve move flips somehow – just that a move that started intraday tends to be stickier than you would think under “normal” circumstances.

You may have noticed other types of “atypical” curve moves this year.  This is not meant to be a complete list, and the markets change all the time.  There is a famous quote that goes, “the definition of insanity is doing the same thing over and over again and expecting different results.”  So at some point, you need to think about what you are doing on this modern curve and constantly assessing what your edge is in the new environment.  Back when I started trading, my manager would say in the 1980s, you could make hundreds of millions of dollars if you knew how to make a financial calculator.  In my day (2000s) you could make millions of dollars just “smoothing” the longer end of the curve.  It is a much harder game now.  Even algos that did well at first start getting crowded out by other algos, until the edges are gone.  You have to constantly adjust and find additional ways of making money in an ever-changing market.  That’s the “fun” of trading!

You probably get some version of that “adjust” chat from your manager from time to time.  The question is HOW do you make adjustments?  Most people have no idea where to start.  Before you go full rogue and start punting crypto, I still think there is a lot of money to be made in STIRS.

Here are some ways you can think about starting to adjust to some of the algo features I mentioned on the previous page:

1.      Find some OTHER ways to generate regular intraday P&L.  Almost all STIR traders focus on smoothing the fly curve.  But there are other ways to generate intraday P&L in the world of STIR futures and options.

2.      If an algo is keeping a part of the curve in-line, think about what it must be doing to maintain that rigidity.  If they keep getting hit (or lifted) on a part of the curve, they must be offsetting that risk somehow.

3.      If an algo or large player is keeping something rigid, think about under what conditions it may release the hold.