Consider the Leverage

Malt Liquidity 120

The other day, I was asked about the performance of levered ETFs and why they aren’t held as part of a diversified portfolio. Other than the simple answers — highlighting the problems with beta decay

If I have an ETF that is levered, say, 2x to the price of crude oil, intuitively this means that if CL moves from 50—>55, a 10% increase, the ETF will move from 10—>12, a 20% increase. But CL is volatile! Suppose that the next day, it reverses its gains and goes from 55—>50, a 9.1% decrease, but ends at the same price as a day ago. The ETF, however, will move from 12—>9.816, an 18.2% decrease. Though the price of the tracked commodity has stayed the same, the ETF has eroded in that span.

or the ability to replicate desired beta efficiently (e.g. holding a basket of stocks that has the requisite beta matching the instrument you want to lever up on; see ARKK) — it’s not exactly clear why this is the case carte blanche, else these products wouldn’t exist (as, depending on what they’re levered on, they don’t just go “straight to zero”). Certainly there is more effective leverage available through LEAPs, but these aren’t exactly intuitive and in any case, you don’t own the underlying with a far ITM option, and it can very easily go OTM (which Qaddafi found out the hard way). But this highlights a pretty topical problem with back-testing strategies in general — how far, exactly, should we look back? Is it indicative?

Take a 5 year period of holding 3X SPY. Sure, you’d be up, but for 3x leverage taken, to just have outperformed the raw index by ~20% seems inefficient. But in a straight bull market like we saw from 2020-2021, you’d have made a killing! Herein lies the inherent issues with “levering to the tits”, as we say colloquially — it always looks great in hindsight. Going back even further, we see an even more notorious outperformance, and perhaps it even made sense to hold this kind of leverage long term in a clearly dead vol environment. In effect, this is what a long-term 3x-levered SPY bet would be — a massive bet on a short vol environment remaining that way. Do you think this is a reasonable thing to assume when comparing 2013-2015’s performance to 2022 onward? Or, in a rising rates environment, would we want to make shorter and shorter time-horizon bets because the dynamics of a frothy rates-rising market in 2023 are very different from the absolute peak of the lending freeze and aftereffects of 2008?

This brings us to the concept of regimes, where the market is fundamentally behaving differently from a previous structure, which could invalidate our ability to backtest and compare entirely. Everyone has their own proprietary definition, but perhaps the easiest way to bucket things is interest rates. Interest rates can function as a barometer of how much risk people are willing to take/are encouraged to take in general. We’re no longer in a low interest rate environment, so naturally, in environments with varying interest rates, we probably shouldn’t treat a backtest period the same way. Trivia — approximately how many times should we apportion out that graph above by this basic, basic definition for regime shift?

On the flip side of things, levered ETFs have a reputation for facilitating degenerate gambling. UWTI/DWTI and, of course, JNUG have long been memes on certain internet circles:

But truly, South Koreans take the cake:

Just in August, South Koreans invested $1.4bn in Direxion’s triple-leveraged semiconductors ETF, making it the single most popular overseas equity investment among locals, according to Korea Securities Depository data obtained and sent over to us by ETFGI’s Deborah Fuhr.

It’s truly delightful that South Korean speculators find it necessary day in, day out to take 3x levered 20-year Treasury positions. I’ve never seen a retail population more addicted to vol for vol’s sake. Just look at this number:

At the end of August there were 1,134 ETFs in South Korea with total assets of $90bn. Leveraged or inverse products accounted for 14.7 per cent of the local industry’s assets — compared to 1.16 per cent for the global ETF industry as a whole

Yeah, if this is how the US ETF allocation worked, we’d nuke to Valhalla within a week. (No, zero-day trading is not going to take us down, as I’ve written about before.)

This seems to be an effect of the new “regime” (on a high level) post-2020 though. With a majority of “actual” money moving to passive, active trades are more and more crowded. Just look at “AI” trades — NVDA first rocketed up on selling GPUs to blockchain companies, and now they rocket up again on selling the pickaxes to mine AI with. Meanwhile, retail traders have essentially decided stock market gambling is fun:

In late June, during a stint in Las Vegas playing in the World Series of Poker, Klett said, he scooped up more than 300 contracts that would pay off if the S&P 500 index rallied by the next day. After he spent a sleepless night checking the futures market for clues on what would happen in the morning, he said, the S&P 500 rose 1.2% and he made $71,000.

He celebrated the big win—and his birthday—by playing roulette and slot machines while hopping from casino to casino on the Las Vegas Strip. “I lost $25,000 in poker but smoked the market,” he said. 

Klett said he places around eight options trades a day, and typically doesn’t hold them for more than seconds or minutes. He has ramped up his activity since losing his job earlier this year, he said, and hopes to make enough trading to buy a million-dollar home.  

Really think about the dynamic of a fundamentals-based actively managed fund. Let’s say I’m running a fund and my research analyst puts me on to some very good bit of financial analysis. The problem is a) the entry sucks because the stock just moves with market beta most of the time, b) guidance after earnings could just bork the trade anyway, and c) nobody actually cares about trading the stock outside of earnings and maybe a couple of days a quarter (no catalyst.) I can’t really blame people for just piling into hedge fund hotels!

While we are not in the full throes of the ZIRP mentality “the dollar doesn’t matter”, the untethering of “company fiscal performance” to “stock performance” is totally caked in at this point. This is why valuations always seem “off” compared to historical levels and why people have been waiting for value to come back when it’s been dead for ages. It’s not growth vs value — it’s passive vs gamblers. Largely there is no reason to do anything else as an individual other than go along with the market (which you can do more efficiently than simply holding a total market index, but I digress), but this is the crux of where financial markets have generally failed in the current regime. It doesn’t make sense for everyone to be invested in baskets rather than actual companies. Something needs to change about market structure in general. But more on that later…