Testing Capital Efficient Gold Assets

Two emails from readers prompt a query: can using 2X Leveraged Gold ETFs improve a portfolio? If so, what to use in the created space? I get behind the wheel to backtest some possibilities. 1st take: 2X Leveraged Gold holds promise; 2nd take: Managed Futures are perhaps the best complement.


Consider this the second installment in a series that I could see myself enjoying quite a bit: Idiot with a Portfolio Backtester (here is the first). These are where I get a question, an inkling, a curiosity, then dial up the Backtest Asset Allocation tool on Portfolio Visualizer, then push a few buttons, turn a few dials, pull a few levers, and then report on my experiments. Don’t take these experiments too literally - I’m not necessarily recommending these portfolios. I just like to put different ideas through their paces and see what I can learn.

This week’s experiment was prompted by emails from two wonderful readers, AZ and JM (thanks to both of you!). They emailed on slightly different topics, but there was some interesting overlap.

AZ wrote about his experiments using UGL, a 2X Leveraged gold fund in his portfolios. AZ noted that the performance for UGL in 2022 was notably worse than for GLD: -6.75% compared to -.4%. Paradoxically, though, AZ found that his portfolio with the leveraged gold fund actually did better in 2022 than it would have with the regular version. By allocating 7.5% to the more capital efficient UGL instead of the normal 15%, he was able to send additional capital to EUO, a Short Euro fund. The version with UGL was superior to GLD in terms of return, volatility, shallower max drawdown and Sharpe ratio.

This was only a study for 2022, though, so immediately I wanted to see how the basic construct might do with a longer timeframe. Also, AZ is fond of using EUO, though these days I have a Schoolboy Crush on Managed Futures, so I wanted to see how using the leveraged gold fund could create more space for that. Additionally, if I were to use a currency fund, these days I would lean towards UUP, a bullish dollar fund (though, admittedly, it amounts to pretty much the same as the short Euro).

With Managed Futures, though, my preferred asset (DBMF - which I gushed over here) only goes back to June 2019. I know there have been managed futures mutual funds over the years, but my understanding has been that these are all pretty idiosyncratic, so they don’t necessarily represent Man Futures in the abstract, or more as DBMF does.

Enter in the email from JM, which suggested using ASFYX, AlphaSimplex Managed Futures Strategy fund, for backtests. It goes back to 2010, which will allow me to lengthen the backtests four-fold. While I don’t love ASFYX in the abstract (expense ratio: 1.47%; one particular version of managed futures, not the comprehensive approach of DBMF), it is great to have it for backtesting. Beggars can’t be choosers, after all.

JM also mentioned that he’s been testing out a sub-portfolio in his longer holdings using UPRO, TMF, and GLD (40/30/30) to get a feel for leveraged ETFs and how they work. Looking at that portfolio, it seems like there could be some benefit from increasing the leverage for gold. If you’re already going with UPRO and TMF, might as well go for it all: in for a penny, in for a pound.

So, then I thought, what if I test out a simple portfolio using leveraged gold funds in different ways to see how they would have done over the past decade plus? And thus, “Idiot with a Portfolio Backtester,” episode 2, was born.

Method

I’ll use the Golden Butterfly as my portfolio chassis - it's simple and straight-forward, with a large (20%) allocation to gold in the original version that gives me a lot of space to play with. I’ll of course include the backtest for this as my baseline.

Then in portfolio #2, I’ll switch out the 20% in unleveraged gold (GLD) for 20% in UGL, the 2X leveraged Gold fund. I’ll keep the rest of the assets unleveraged and at 20%, so we can see what an outsized position to gold would do to an otherwise balanced and fairly conservative portfolio. This means that the portfolio is a 40/40/40.

For portfolio #3, I’ll also use UGL, but this time at a smaller allocation and then use the extra space created for higher nominal allocations in the other four assets. I’ll put the allocation to UGL at 12%, and then add 2% each (nominally) to the other four assets. Although there will be a slightly bigger bump to gold (24%), it still keeps the five components of the portfolio roughly in line.

For portfolio #4, I’ll keep the original 20% for the four standard asset classes, but 10% into UGL and then use the remaining 10% to send to the Managed Futures fund, ASFYX.

Portfolio #5 will also have the 10% to UGL, with the other 10% going to EUO, AZ’s choice. The benefit here is that we’ll get a twelve year backtest.

Finally, portfolio #6 will use UPP, the Bullish Dollar fund, instead of EUO. I’m expecting that performance will be near identical to #5, but we’ll see. There’s also an interesting conundrum by devoting some money to a bullish dollar in the sense that it is, conceptually speaking, the flip side of strengthening gold prices. Is this, or even portfolio #5, just a way of canceling out your gold position? Hmmm….

I’ll use the “rebalance bands” setting, in keeping with my recent decision to use rebalancing bands in all my portfolios and some resources I have found over the past year.

Results

Here are the links to the backtests: this one for portfolios #1, 2 and 3; and then this one for #4, 5 and 6. The start date for all backtests is August, 2010 (constrained by ASFYX).

Credit to Portfolio Visualizer

Some of my observations:

  1. The best portfolio, hands down, is portfolio #4, where the space created by allocation to UGL was used for managed futures. It is the highest in CAGR and Perpetual Withdrawal Rate, my go-to indicator for assessing portfolios. Not that I pay too much attention to Maximum Drawdown, but its performance here was notable, as well, as it was about 3.5% less bad than the Classic Golden Butterfly and the best of the five portfolios. This portfolio was second-highest for standard deviation, but then again, a 7.74% standard deviation is very stable by any reasonable standard.
  2. Portfolio #3, which lowered the allocation to gold (since the use of UGL meant it had greater exposure to it) and used the space created by the leverage to distribute to the existing other asset classes, outperformed #2, which just had way more exposure to gold. In AZ’s example, he used UGL in a mix with other leveraged assets which may be smart; here, we are using UGL as the only leveraged asset, which appears not to be a great idea.
  3. Then again, the Sharpe Ratios for all “UGL & something else” portfolios were all about the same as each other and all better than portfolios #2 and 3, which kept the same five parts as the Classic Golden Butterfly save for the use of Leveraged Gold instead of regular Gold. This suggests that capital efficiency with the gold allocation can yield positive results, if you use the extra space for some other less correlated asset class with a positive return.
  4. Between the two currency funds, #5 with EUO was better than #6 with UUP, but the performance of each was pretty dull. They were more stable portfolios than even the Classic GB, but involved a sacrifice of returns, as well. For #5, you have a pretty good Sharpe Ratio, but you would have sacrificed 45 basis points of growth for a few extra basis points of Sharpe performance, which doesn’t seem like the best trade.
  5. Since #5 and 6 ended up about the same, and made for more stable portfolios. It does appear that the allocation to short Euros or long Dollars had the effect of counteracting the gold fund, as I wondered about. This is a bit outside of my expertise (must admit: thinking through currency plays always confuses me), but is a data point to put a pin in for the future.

Conclusions

Summing it up: Count this as a win for using Leveraged Gold in a portfolio, but I’m far, far, far from having fully formed thoughts on the use of Embedded Leverage ETFs like UGL in portfolios. I know, I mention leveraged ETFs like UGL, UPRO, DRN, TMF, and others once in a while in the blog, but I want to make sure I’m being responsible with how I talk about them. I’m not quite there yet to say that Leveraged ETFs are a good and responsible choice for the average investor.

Here, though, the capital efficiency of getting the same exposure to gold through a smaller dedicated allocation in real terms led to measurable improvement in portfolio performance from 2010 until now. There was nothing special going on for Gold during this time period. It didn’t even do that well, but using leverage to get more of it led to better results.

Using that capital efficiency to get exposure to Managed Futures was the best option of those that I tested. It seems like every time I ask a question, Managed Futures seems to be the answer. What heartens me even more is the arrival of new Managed Futures ETF like DBMF, and even CTA and KMLM, with lower expense ratios. Good times.


💡
Disclaimers: I own shares of some of the ETFs mentioned in this article, such as DBMF, UPRO, DRN, TMF and GLD, though I don’t own any shares of the main character of the story, UGL. Keep in mind that this site is intended for financial education, and this write-up is to explain how you might want to think about particular assets. It's not a recommendation that you invest this way in real life.