Shorting The Short

David Orr who runs the Militia Long/Short Equity ETF (ORR) that we took a quick look at last week Tweeted out a stream of consciousness type of list that had some interesting opinions. The one that might be most relevant to the conversation we have here as follows.

Sample size is underappreciated. Any track record relying on 10 or fewer bets to have beaten the market is no track record at all.

Moreso than tackle that directly, I think there is an interesting offshoot to this point. We talk about this some but when you try to understand how well a holding has done or how well a portfolio has done, it is important to understand the track record. I don't mean just see and digest the numbers but when something looks eyepoppingly good, too good, I would try to understand why it was so good, or bad maybe, and try to assess whether that result has any reasonable chance of being repeatable. 

The Invenomic Fund (BIVIX) that we've looked at a few times is seriously skewed by two phenomenal years in 2021 and 2022 when it was up 61% and then 50%. There's nothing horribly wrong with the other years although the track record does look kind of short. I asked three different AIs why it did so well in 2021. If I had gotten a real answer I would have asked the same for 2022. The best answer of the three was pretty vague "The Invenomic Fund’s strong performance in 2021 can be attributed to its value-biased long/short equity strategy, which capitalized on the market’s shift toward value stocks, its disciplined and diversified investment process, and the experienced management of Ali Motamed. The fund’s ability to navigate sector rotations, leverage short selling, and adapt to a volatile market environment likely drove its significant outperformance, particularly for the Institutional share class (BIVIX)."

I think I was expecting AI to be able to find old commentaries and annual reports from that time but no luck. 

The bigger thing is not banking on some sort of great performance that can't be repeated. I think that sort of mindset can help minimize the extent to which investors chase heat. Things like volatility and correlation are more likely to be repeatable than some sort of scorching hot return. 

This might sound contradictory to the idea of certain sectors or industries regularly outperforming. Per testfol.io, tech as measured by simulated XLK has outperformed the S&P 500 by 134 basis points annually for the last 30 years. Tech outperformed 17 out of 31 full and partial years in that period and in nine of those 17, tech outperformed by more than 10% and in five of those nine, tech outperformed by more than 20%. 

The flipside is tech tends to go down more on the way down. Tech is a major sector with reliable tendencies versus a specialized strategy that relies on models being right. Those are different things. 

Circling back to Militia Long/Short Equity ETF, one of the things that intrigues me and that I believe is unique is that the fund shorts ETF structural inefficiencies like volatility drag and the erosion of the crazy high "yielding" derivative income ETFs. It also looks like it is currently short some business development companies (BDCs).

I am aware of one fund filing for a fund by Defiance that will short inverse and levered long MSTR ETFs. I mentioned that David Einhorn does a variation of this trade where he hedges that sort of MSTR position by going long the common. We backtested the idea and it works. I think there is something to this trade, that it can succeed (the hedge version that Einhorn reportedly does) and we may see more funds putting this trade on. 

So let try it with a 5% short position in SPXU which is the ProShares 3x Short S&P 500. In 17 full and partial years, it has only been up twice. It's down 99% in its lifetime. The history of these funds is they go down a bunch and then reverse split. SPXU has reverse split five times. To be clear, the following portfolios a short a 3x inverse fund. 

It doesn't really show up in the backtest but 2022 was the only year in the short period studied where shorting SPXU worked out badly. 

The volatility goes up some versus VBAIX but the gain in return seems to be worth it at least as far as the Sharpe Ratio is concerned. If we take out Portfolio 2 to get a longer look, we can see back to 2013 and the stats again look pretty good versus VBAIX.

Part of the blogging process is to circle back to previous topics to see if something has changed or maybe something changed for the better or worse. Over the weekend I thought about the Ultra Risk Parity ETF (UPAR). Risk parity generally leverages up on bonds to even out the risk allocation of a multi asset portfolio. RPAR is an unlevered risk parity fund and UPAR is 2x RPAR. I've never been a fan of either fund but they are interesting. Coincidentally, ETF.com included UPAR in a Tweet on Monday about alts that are doing well in 2025.

I think we've only looked at UPAR as a standalone holding, similar to how we look at Vanguard Balanced Index Fund (VBAIX). Maybe it is better thought of as a capital efficient piece of a portfolio instead. I built out the following, comparing the same portfolios with one using UPAR and the other using AQR Multi Asset (AQRIX) which used to be AQR Risk Parity. AQRIX still uses leverage in a manner that I'd say resembles risk parity. 

In 2022, UPAR was down 30%. Bonds got hammered of course and UPAR had 98% in bonds with duration plus some equity exposure. Despite that hit though, Portfolio 1 was only down 9% because it was a good year for SPMO and managed futures. Portfolio 2 was down less than 1%, AQRIX alone was only down 10%. 

For the limited time we can study because UPAR only goes back to early 2022, Portfolio 1 was able to keep up with inflation despite UPAR compounding negatively to the tune of 7%. The conclusion is not to buy UPAR, even in a smaller weighting but Portfolio 1 overcame UPAR's dreadful returns to still keep up with inflation. UPAR is working this year which is fine for anyone interested but the bigger takeaway is that we can be patient when something with a 5-10% portfolio weighting is struggling. 

I have not been tempted to throw in the towel on managed futures for example. It has been struggling for a bit of course although things may have turned a corner a couple of weeks ago, we'll see. 

When we take Portfolio 2 back to when AQRIX changed its name in early 2019, the portfolio outperformed VBAIX (now were talking just over six years of data) by 123 basis points annually with considerably less volatility. In 2022, Portfolio 2 was up less than 1% as I mentioned versus down 16.87% for VBAIX. 

The information, analysis and opinions expressed herein reflect our judgment and opinions as of the date of writing and are subject to change at any time without notice. They are not intended to constitute legal, tax, securities or investment advice or a recommended course of action in any given situation.