Avoiding Overly Sophisticated Portfolios

Let's continue the conversation about all-weather generically and then the Cockroach Portfolio.

First a comparison of the Permanent Portfolio Mutual Fund (PRPFX) versus a 60/40 portfolio comprised of two Vanguard mutual funds. We can test this all the way back to 1993.

PRPFX has a lower standard deviation and it's market correlation is 0.67. You can see over the 30+ years it has lagged by 114 basis points annually. This period is noteworthy.

During the five years of the internet bubble, PRPFX lagged 60/40 by 12%-17% per year which would have been very difficult to endure. Going from 2003, forward which strips out the internet bubble and recovery, PRPFX still lagged by 60 basis points annually and surprisingly, had a higher standard deviation. PRPFX outperformed 60/40 in 2008 by almost 12 percentage points and in 2022 by 11 1/2 percentage points. For the last ten years, 60/40 has blown away PRPFX and the standard deviation numbers for PRPFX in that time were only so so.

In February we built a do it yourself version of the Cockroach Portfolio which is a "derivation of the permanent portfolio" as follows. The funds chosen, allow us to backtest for ten years.

And the results versus the same 60/40.

Back in February, we also backtested it without Bitcoin, and the growth was only 33 basis points annualized. 

Putting 20% in VIXM pretty much doomed that version of the portfolio. The actual Cockroach's allocation to volatility is dramatically more sophisticated than what I did. 

I think 20% into volatility, 20% into managed futures and then 16% into gold is way too much into assets that you don't expect to look like the stock market. Stocks are the thing that go up the most, most of the time. Protection against some of the equity market's volatility makes sense to me but if you neutralize it entirely, you won't get much growth along the lines of DIY Cockroach without Bitcoin. 

Either the Cockroach can't be easily recreated with retail accessible funds (seems more likely) or it is too conservative to grow (seems less likely). That doesn't mean we can't take a bit of its process to inject into our portfolios. Arguably, I've done this with small allocations to volatility strategies. I'd been using managed futures, which Cockroach is big on, since long before I ever heard of the Cockroach.

Circling back to the idea from the start of my blogging days that funds will get more and more sophisticated, democratizing access to strategies that were previously unavailable, I think we can use a couple of newer volatility funds that might introduce better results into today's discussion.

The backtest is ridiculously short but we're working on portfolio theory. Here you can see the makeup of the Cockroach as we mimicked it in February and Cockroach Light which has a much less in volatility funds and 60/40 is Portfolio 3.

We've talked a lot about CAOS and a couple of times about MAXI. CAOS pairs BOX spreads as a cash proxy with a put option overlay. I think they have the cash proxy part of the fund very dialed in but I am not convinced the fund will offer too much tail risk protection, I am curious to see if it ever will. I mentioned MAXI the other day. It owns Bitcoin and it sells option combos on broad and narrow index ETFs. The options strategy should not get in the way of whatever bitcoin does as opposed to the Roundhill Bitcoin Covered Call ETF (YBTC) where the options are bitcoin options which will peg the price of bitcoin and limit gains if/when bitcoin moves up quickly. 

Where we talk about long volatility strategies and short volatility strategies, MAXI appears to be a long/short volatility fund. Owning Bitcoin could be thought of as long volatility and selling options combos is short volatility. MAXI is bitcoin via futures and income from the options strategy. If bitcoin turns out to be a complete scam then MAXI should be expected to go to zero. A percent or two into something that goes to zero is not going to blow up a financial plan even if it would be very disappointing.

That DIY Cockroach, our attempt to mimic the real thing, did so poorly even with 4% bitcoin is surprising. The huge weight to VIXM is probably what did the portfolio in. Cockroach Light outperformed 60/40 by a noticeable amount as a matter of luck probably thanks to bitcoin and the standard deviation is much lower probably due the very large weightings in RYMFX and client holding BKLN. 

In dissecting Cockroach Light, I am not worried about a broad based, equity index fund malfunctioning or somehow breaking. The threat is that occasionally, it will go down a lot. Bank loan funds are mostly lower risk but the space did blow up in 2008. I found two mutual funds that were around back then, EVBLX and LFRAX, that were down 30% and 21% respectively that year. I would not put 30% into bank loans IRL but I've had a small allocation for many years. T-bills instead would be less volatile with a lower yield. Managed futures can be a very challenging hold as we've talked about countless times. The group struggled for many years in the 2010's and we saw a nasty whipsaw a year ago. That whipsaw did not truly break the strategy but was a good humbling for anyone who actually put 20% or more into managed futures off of 2022 hype. BOX spreads used in CAOS are a legitimate cash proxy so as I mentioned that fund's biggest threat IMO is turning out to be an ineffective hedge with a very modest bleed from puts expiring worthless over and over. And finally MAXI which I mentioned relies on bitcoin turning out to be a real thing. 

The category weightings work but I think doing anything along these lines would require more holdings for fixed income and dividing the managed futures sleeve into several different strategies.  

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.