Leveraged ETFs are not for long term holding, and the main reason is that the loss and recovery returns are asymmetric, coupled with the fact that large drawdowns do occur in the market regularly.
The below chart shows the profit from investing 10k in SPXL in red, and 30k in SPY in blue, with the green line showing the advantage of SPXL. No doubt the returns are better (since SPXL only uses 10k) but the total profits are actually better with 30k in SPY, with lower drawdowns.
|Initial PV||Final PV||Profit||Return||Max DD||Max DD Value|
The below chart shows the profit (dollar) differences between the 2 portfolios mentioned above. Increasing the time horizon makes the average return more negative, but fattening the tails to the right as well, indicating that a well-timed mid-term (2-4 years) investment in SPXL can actually beat the SPY by a large margin, but in most cases you will lose out.
As you extend the horizon, the SPXL becomes more like an OTM call option on the SPY. You're more likely to lose, but are also more likely to have an outsized gain.
Asymmetry in required returns for recovery
The asymmetry between percentage losses vs percentage gains required to recover makes a portfolio increasingly difficult to recover from larger drawdowns.
This, coupled with the nature of average daily returns and the presence of large drawdowns, combine to make leveraged ETFs unsuitable for long term holding.
Nature of returns
Ultimately it's the volatility that kills the leveraged investment PNL in the long term. The longer you hold it, the more damage it can do.
By resampling, we can find the breakeven volatility required for an ETF to guarantee a better return than the non-leveraged counterpart.
The below charts show portfolio profits using resampled returns from fitted t-dist on SPY returns, and simulated 10000 trajectories, each over 250 days. Note the cutoff at the -10k and the -30k mark - these are when portfolios get completely wiped out.
Chart 1: fitted t-dist, showing that the leveraged portfolio has a higher mean, but lower median return. On average it does better, but more times than not, it does not.
Chart 2: reducing the mean return by 50%, the comparison still stands.
Chart 3: reducing volatility by 50% really makes the leveraged portfolio shine. The less the volatility, the more certain it is to outperform.
Chart 4: increasing volatility by 100% shows that volatility is really the key in making or breaking the outperformance. Here you can see that the leveraged portfolio payoff distribution resembels an OTM call option - it is very likely to 'expire worthless' in that the entire 10k is lost.
The sampling is not entirely representative to actual markets, because there are other things to consider like autocorrelation - the simulations above only give IID returns. But it serves to identify in which cases the leveraged portfolio outperforms - in this case we know it's when volatility is lower.
When does it pay off to hold the leveraged SPXL then?
When volatility is expected to be low.
From the previous analysis, we've seen that even with a low mean, a lower volatilty increases the likelihood of outperformance. Since mean returns are not expected to be negative in the long run, the best time to (if ever) change to a leveraged investment is when you are predicting a stable (and ideally rising) market.
Since picking the right time is not easy to do (and certainly the environment now does not warrant such a view), it's generally better to have an unleveraged portfolio - if cash is the concern, use other means, e.g. instead of 1x SPY, you can consider using options to replicate it.
Not only do you have to pick the right time to get in, you also need to pick the right time to get out, otherwise your gains can evaporate really quickly. Here are some historical performance comparisons over some periods.
This illustrates how a drawdown can set you back so bad that the following rally is not able to claw back the underperformance. Note how the leveraged portfolio did 'better' in the drawdown - this is because it started with a capital base of 10k so it was actually severely depleted, thus further losses aren't as painful in dollar terms - but this is the very same reason why it could not capitalize on the rally.
This shows how even a rally pockmarked with even light to medium severity drawdowns can cause the leverarged portfolio to underperform.
This shows how quickly the outperformance can evaporate on a medium severity drawdown. It's psychologically uncomfortable - you're amplifying your emotions for nothing.
Notice how the red only caught up with blue when the rally was really strong. Then again it quickly gave up its outperformance.
Again showing how quickly outperformance evaporates and turns into underperformance.
- It's recommended not to use leveraged ETFs for long term investment - it's better for speculating over short term horizons - arguably you can/should use calls.
- The outperformance of leveraged ETFs can quickly evaporate - not good for psychological well-being, which is important to long term investment success.