Submitted by Peter Tchir of TF Market Advisors,
I have two major concerns right not that I think everyone should be nervous about. Actually I have lots of concerns, but most aren’t about the markets, and discussing them in this forum isn’t appropriate, so I will stick to my market related fears.
These are two risks that I see taking this sell-off further and faster than anyone else expects.
The Death of the Normal Curve
I think algo’s in general have had changes the distribution of returns. That seems particularly true around the big Moving Averages. Maybe it is just me, but I see so many conversations about the 50 DMA and 100 DMA these days that it seems that everyone is looking at them. It might just be my streams, or it might just be that everyone really is that bored, or is looking for an excuse not to sell.
In any case, I think there are many more “entities” trading the moving averages (and other technical levels). I think that has shifted the return distribution over time.
Let’s say in the good old days, there was a “normal” return distribution around the 100 DMA. Since I am graphically challenged I will work with a very simple binomial distribution that illustrates the point.
Say there was a 75% chance of a bounce with a 0.5% profit, then there should have been a 25% chance with a 1.5% loss. That would have created an expected value of 0.
People still trade it because they felt they had some other “advantage” that let them pick the bounces with a higher success rate. Or pick those bounces that would generate higher returns.
Over time as more and more algos try to trade the same phenomena the success rate actually increases.
It becomes to some degree, a self-fulfilling prophecy. If I am buying at the 100 DMA because it bounces and you are buying at the same price (for the same reason) then it is likely to bounce. So what happens over time is the “win rate” increases.
Now let’s say you win 90% of the time, and the average profit is still 0.5% then that loss, which only occurs 10% of the time, should be 4.5%. Yikes.
You can see how it would happen. The 100 DMA at one time represented the entire market which was not overly biased towards technicians. Shorts maybe covered at those levels because it seemed appropriate. Longs doubled down because they had liked it there once before, so why not now. All these charts are just a graphical representation of human behavior.
But now that has changed. A lot of the people sitting on long positions bought for no other reason than it should go higher. It wasn’t a long only manager adding to a position at levels they had once liked. It was a twitchy algo buying because it has to go up. One feature all those algo’s have is relatively tight stop losses. They may all need to exit at the same time. That might push us further than anyone expected.
While real money might add at the 100 DMA, maybe once they break through it and have all their gains from the past few months wiped out, they don’t add, or even sell. Maybe the algo’s that sniff out weakness short. Maybe that is why we don’t get a small gap down, but instead hit an air pocket.
So I am nervous that the support we think we have has been eroded by the types of trading that goes on, and that what should be a small sell-off based on the data, becomes a larger sell-off base on the positioning and types of trading we see in the market.
In case you missed it this morning, we recommended covering the long bond long position (I do love saying that).
Before getting into why we are nervous, I have to admit I still think TBT might be the most insane “investment” out there. 2 times the daily move in any treasury seems silly. These leveraged ETF’s have serious path dependency problems as it stands. The “churn” of daily bounces hurts their returns. That is common enough in treasuries. Then it is based on some index (Barclay’s 20+ year treasuries) that is completely affected by the Fed’s positions. The Fed owns significant portions of a lot of the bonds that are in that index, making pricing less transparent. But that is solved by being short through bilateral swaps. Okay, I use the term “solved” very loosely. So you have path dependent leveraged risk to an index that isn’t fungible through bilateral swaps. Shoot me. Please just shoot me.
But shares outstanding for TBT continue to grow. If ever any investment should be destroyed on principle alone, this is it. It makes me want to go long the long bond, again.
But I can’t. Not every “consensus” trade loses every day and I am very nervous that “something” is going on in the treasury complex.
Here are our concerns:
We have hit some target levels 2.60% on the 10 year and 3.55% on the long bond (or close enough)
We aren’t rallying as much as we “should” be. Completely subjective, but this morning when futures were down 4 (how quaint that seems) treasuries were also lower. It feels like there is real resistance here, and I’m not sure how things like TBT aren’t causing a massive short squeeze, but it appears that they aren’t. So cautious here.
The prices paid on ISM was very high today. The CRB index is getting higher by the day. The dollar is weak. So there are signs of commodity inflation, if nothing else, but that should help put a floor on how low treasury yields go.
Finally, and possibly most important, is that we are annoying most of the rest of the world. Bernanke often said that trade barriers established during the great depression made the problem worse. He urged government not to repeat that act, and they haven’t. But his policies are starting to have that same impact. Countries blame QE for their mess. Countries are starting to question the sense of having a single reserve currency.
So I don’t like what is going on here.
What is worse, is that I do believe that if treasuries crack at all, then retail will exit high yield and investment grade bonds and with spreads already leaking there will be no hedge fund demand. In fact, hedge funds will become forced sellers. Then the real final bid, the pension funds and insurance companies, who are already fairly long risk, will hold off using their capital until the pain grows.
It will be the drop in corporate bond prices that cause the real problem, but it will be precipitated by a treasury sell-off. One that we haven’t seen yet. Hopefully we won’t see. Hopefully some perception of “safety” and concerns about how weak growth will be will hold down treasury yields, but I am extremely nervous.
We are selling the remaining 5% of SPY March 180 puts. We want to sell them before they become completely intrinsic value.
We should probably sell XOVER protection, but will hold on to it for now (in no small part because it is now after noon and London has shut).
We might take some IG off later today (we added a tiny bit more on Friday around lunchtime).
I am very close to going short treasuries. Not the long bond. Maybe the 10 year. Maybe the 5 year. Maybe we do via interest rate swaps. I am looking to see what expresses my concern the most with the least amount of damage if I am wrong. This is difficult to pull the trigger on since you know my feeling on TBT and since I am bearish enough stocks that it is hard to get too jazzed up about a treasury sell-off.