An Online Origin Story: Hedgefundie's Excellent Adventure

Way back in 2019 - before Wall Street Bets was mainstream - I stumbled on a slightly out of place thread in the investing forum, Bogleheads.org. Bogleheads are self-described disciples of Jack Bogle, the founder of Vanguard, the leader in low-cost, passive index-based investing. So I was a bit surprised to see that one of the most popular threads on the forum was kicked off by the poster, Hedgefundie, describing a strategy using 3x leveraged ETFs to replicate an aggressive risk parity portfolio. After thousands of replies and many iterations, Hedgefundie announced that his official recommended portfolio was as follows:‍

• 55% UPRO

  (ETF seeking 3x daily returns of S&P 500)

• 45% TMF

  (ETF seeking 3x daily returns of long duration treasuries)

Much of the inspiration for this strategy originates from a whitepaper written by Bridgewater’s Ray Dalio titled Engineering Targeted Returns and Risk. The most influential point in the paper is that low-risk, modest-return assets can be leveraged to provide higher returns (with commensurately higher risk). When uncorrelated, leveraged assets are combined in a diversified way, the resulting portfolio can provide higher risk adjusted returns than a traditional, unlevered portfolio. Without leverage, a traditional portfolio must make the tradeoff between overweighting equities to hit target returns (and thus being undiversified) or accepting low returns for the sake of broad diversification.

In practice, Dalio’s proposed solution is referred to as a “risk parity” strategy. The following diagram from the whitepaper illustrates how a risk parity portfolio differs from a traditional portfolio:

While the Boglehead’s attempt at replicating Bridgwater’s risk parity fund is an incredibly simplified version of the real thing, the popularity of the thread was illuminating. After staying at the top of the Bogleheads forum for over a year, other retail investing forums started to pick up on the approach - including the infamous Wall Street Bets. Clearly, the eye-popping returns were captivating a new audience that was unfamiliar with a strategy that had long been mainstream amongst hedge funds and institutional investors. In the next section, we will inspect Hedgefundie’s strategy using Composer and try to determine if the enthusiasm is justified.

Breakdown of the original Excellent Adventure

Given the simplicity of Hedgefundie’s 55/45 blend of UPRO and TMF, it’s easy to create and backtest this strategy in Composer. We define the strategy as follows in the symphony editor:

Data for the leveraged ETFs only goes back to July 2009, so we use that as the start date for our backtest:

The green line represents Hedgefundie’s Excellent Adventure (HFEA), while the yellow line represents UPRO and the red line represents TMF. This way we can compare the strategy to its component ETFs.

On the plus side, the blended strategy certainly outperforms its constituent ETFs in risk adjusted returns, besting both UPRO and TMF in terms of Sharpe and Calmar ratios. Perhaps most importantly, the strategy isn’t particularly strongly correlated with either ETF, and its sensitivity to both benchmarks is quite modest. The surprisingly good performance when blending the two ETFs can be explained by the negative correlation between the two assets, with a Pearson correlation coefficient of -.45 during the backtest period. In layman's terms, when UPRO zigs, TMF zags.‍

Still, there are some issues with HFEA. Even if it has a muted, 44.2% drawdown compared to the gut wrenching, 76.8% drawdown of UPRO, it’s still quite a bit bigger drop than SPY, which “only” has a 33.7% drawdown during the backtest period. But perhaps most worrying is that the correlation between TMF and UPRO isn’t stationary - there is no guarantee that the correlation between the two assets will remain negative, which could erode the volatility-reducing properties of blending the two ETFs. And this is less of a hypothetical than a reality, given the sell-off in everything during the start of COVID, triggered in part by risk parity funds reducing risk exposures across the board. We can see a very pronounced drop in HFEA, TMF and UPRO  during the COVID crash in March 2020, with even TMF suffering a 43.8% drawdown - a phenomenon known as “crash beta,” when all risk assets drop at the same time:

A More Refined Adventure

For those still eager to embark on this adventure, some simple modifications to HFEA’s strategy can help you weather — or altogether avoid — the inevitable storms. Using Composer’s no-code symphony editor, we can sand some of the sharp edges off of HFEA, a strategy we title “Hedgefundie’s Excellent Adventure Refined”:

In sum, if we’re in a risk-on environment, we don’t change anything from HFEA’s original implementation shared on Bogleheads. But if we think we’re in a risk-off setting, we want to make sure we don’t get hit by crash beta. And to be more responsive, we run our symphony every single day, as opposed to quarterly.

The yellow line is HFEA’s original strategy, while the green line is our modified, refined version. Now, obviously the return of the yellow line is more insane than the green line — but keep in mind that the purple line represents the SPY! To put the performance in terms of dollars and cents, $100,000 invested in HFEA Refined on July 4, 2009 would be worth $2.08 million as of August 10, 2021. By contrast, $100,000 invested in SPY over the same period would “only” be worth $619,600.

More to the point, our refined version has a much lower max drawdown - lower, in fact, than even the SPY:

While the Sharpe Ratio is similar to the original, the refined strategy has a considerably higher Calmar ratio.

If we zoom in on the COVID crash of 2020, we see that the “crash beta” all but disappears in our refined version, as evidenced by a mere 12.4% drawdown (vs a ~40% drawdown for SPY):

Future Adventures

The symphony’s backtests above are encouraging. But, as everyone knows, past performance is not an indicator of future successes. In addition to running backtests using historical data, Composer makes it simple to follow a symphony’s performance in real-time. Following a symphony is free and carries no risk.

To follow Hedgefundie’s Excellent Adventure Refined, click on the link below to see the symphony details, create a Composer account, and click “Follow”. There’s no better way to understand a strategy than to follow it and see how it responds to changing market conditions. And if you like what you see, it can help give you confidence to invest.

Check out Hedgefundie’s Excellent Adventure Refined