My impression is that market efficiency varies a lot from year to year.

Recessions tend to be associated with less efficient markets. I presume that's because it takes a lot more capital to correct mispricing when the overall market makes large moves.

If a phenomenon only exists one year per decade, then it's less valuable to have enough liquid capital to exploit it, fewer people have the patience needed to remain alert enough to notice it, etc.

2020 has had mispricings that seem more unusual than once per decade phenomena.

The US presidential election has already ended but for weeks since it concluded people outside the USA (and arguably inside the USA) have been able to bet on it. At this point, it is very clear that Trump will not become president. But you can still make 20%+ returns shorting 'TRUMPFEB' on FTX. As recently as the 24th you could make 30%+. These markets resolve on Feb 1st. This easily 'beats the market' even if you give Trump a 1% chance and accept some counter-party risk.

I can think of various other ways to easily get 10%+ returns in months in the crytpo markets. For example several crypto futures are extremely underpriced relative to the underlying coin. At this point, I am wondering why the EMH was ever held in such high regard on lesswrong. These markets have fairly high liquidity. You could have gotten in millions of dollars. 

One common response is "If you are right why are you not amazingly rich?". This seems to assume that being right means you can convince people to give you tons of capital to work with. It is unclear why 'being right' would get you rewarded in this way. Notably even if you post advice publicly and it works you do not magically get billions of dollars. For example, I told people to Bet on Biden [LW · GW]. Joe Biden did win but I did not make millions off this. Grey enlightenment has been posting very good financial advice for years and is likely quite wealthy. But I predict he is not a billionaire and does not run a hedge fund. His long and very good track record does not magically cause him to have access to billions in capital to invest.

Thanks for the post. Here is Vaniver's comment [LW(p) · GW(p)] arguing against timing markets that I really liked, but it was mostly using it as an analogy for when to get out of countries, because there isn't a big cash prize if you time it right, and if you're wrong you get killed or something.

I've been saying this for years.  EMH is just sour grapes, it is exactly like all those news stories about how people who won the lottery don't enjoy their money.

Whenever there is a thing that people can do, and some don't, a demand exists for stories that tell them that they are wise, even heroic, for not doing the thing.  Arguments are Soldiers, Beware One Sided Tradeoffs, all those articles sort of gesture at this.  That demand will be met because making up a lie is easy and people like upvotes.

EMH is a complicated way to say 'your decision to do nothing was the best one.', even when that manifestly isn't true.  Try and write down what people will say before them 'I make 70-200% without risk in 2 months' and see if you get bingo.  'The House Always Wins' is your free middle square.

Help me time the SPAC bubble

My basic approach with bubbles is balancing. Think in terms of risk exposures rather than entries and exits.

Consider how far the price might drop when the bubble finally pops (sometimes that's to zero) and consider how many dollars you're willing to lose on the bet. So, in the case of something that could drop to zero, that means never leaving any more than your maximum loss in dollars in the investment. As the bubble inflates, sell off shares to stay below that value. That means you're locking in gains. When it does finally pop, most of what's left at risk in there might evaporate, but you've kept your profits on the way up. If you think the floor is some value higher than zero, then you risk a fixed amount above the floor.

If it dips significantly, it might be about to pop, or it might just be a swing on its way up. Try to keep your risk constant by putting more dollars in. This can be hard. Be systematic rather than emotional. E.g. maybe pre-commit to putting more dollars in once your investment's value drops below 90% of your max risk. But don't trade too often or you'll burn too much in transaction costs. The exact details of your system are less important than actually using a system.

You might run into granularity issues if the price of a single share is above your risk target. At that point, you're done. Sell your last share and get out.

This is pretty much the strategy you'd get by Kelly betting (or fractional Kelly betting) on a strategy that's a small part of your portfolio. For multiple simultaneous strategies that are a significant fraction of your portfolio, Kelly will also make you balance strategies against each other instead of just against cash.

Can we do any better than this?

Unless you've analyzed a real anomaly that lets you predict the crash better than the market, then the answer is probably "No." Although, if there are derivatives like options or futures or correlated assets, then maybe you have more (heh) options on how to trade this.

But there's so much noise in the markets that it takes a lot of data to try and predict things better than the current price. A SPAC that only lasts two years is not going to have enough price history. But sometimes price anomalies cut across entire sectors. If you can aggregate returns data from multiple SPACs, then maybe you can try to analyze factors that apply to all of them (or most of them). The basic process is like I described in Charting Is Mostly Supersition. [LW · GW] Sometimes you can find an anomaly that lets you predict things a little better than chance. Even small edges can be very valuable.

I would worry in a lot of these cases that there's some risk that your model isn't taking account of, so you could be "picking up pennies in front of a steamroller". Not in all cases though - 70-200% isn't pennies.

But things like supposedly equivalent assets that used to be closely priced now diverging seems highly suspicious.

Two securities (symbols to come later as this is still being actively traded)

Are you trading at a scale where this is relevant? The vast majority of symbols have volumes where a mere mortal's bank account won't move the market at all.

Great post overall! There seems to be a lot of recent interest on LW around playing the markets well. I'd love to set up a Discord for ongoing discussion and brainstorming, if there's interest.

I've been sitting on a good nugget of crypto earnings for years and have had a lot of analysis paralysis about diversifying into the real markets and/or arbitraging the highly volatile crypto options/futures markets directly. If anyone reading this is interested in a serious discussion, please reach out!

These SPAC options have really high implied volatility. Makes me want to sell them. The market tends to overestimate option volatility around known events because demand for options is very high at those times. Maybe a SPAC acquisition is similar. I'm not sure.

A covered call, or equivalently, a deep-in-the-money cash-covered put seems like a pretty good bet. The puts are simpler since you don't have to buy the shares, but the calls probably have better liquidity. If the bubble keeps inflating, you can keep some of that upside. If it pops, the premium from the option makes it hurt less. And you don't expect the shares to end up worth less than $10, which limits the risk even more.

Out-of-the-money short puts also look promising. Between the premium and the $10 floor, the 12.5 May puts on both HCAC and TRNE look like free money right now. If it does end at $10 in May, you get nothing, but if there's a vol crush or price spike between now and then, you could exit and collect most of it early.

I'm also wondering about exotic spreads between the options and the warrants. The warrants seem a lot cheaper than the options. +27 HCACW / -1 HCAC 17.5 call (any date) also looks like free money right now.

[Epistemic status: thinking aloud about how else I might trade this. Not investment advice. There may be risks I'm not seeing yet. Don't bet the farm.]

Market efficiency doesn’t come from the magic all-knowing Omega, it comes from a lot of traders looking for inefficiencies, exploiting them and thus closing them. Broadly speaking, there are 2 ways to do it: manually and algorithmically.

Manual trading is slow. A person has to find some information, reason about it and make the trade. This happens at the days/weeks/months scale, so for some non-negligible time the opportunity persist. If you have the proper mindset that allows noticing inefficiencies like those in the examples and you happen to notice them first, you can exploit them.

Only algorithmic trading allows to react instantaneously and remove inefficiencies beyond millisecond scale. However, it is not flexible. You need a big upfront investment in time and money to notice the inefficiency, develop and backtest your strategy, etc. This only makes sense for regular, systematic inefficiencies.  AHT and HTZ examples just cannot be handled by algo trading.

Equivalent asset arbitrage (and stat arb) is a perfect fit for algo trading. However, we still sometimes see huge inefficiencies there. How could it happen? The arbitragers have risk limits and these risk limits can get overwhelmed by market dynamics. Suppose there are 2 assets A and B that are either formally the same (as in the Wei Dai’s example) or are almost the same (e.g. BTC swap and spot prices). There are professional arbitragers with some total cumulative capital. When the spread diverges, they bet on the convergence and by doing so make it converge. Normally, everything is ok and the spread is minimal and almost unexploitable. However, under extreme market conditions (e.g. LTCM or Bitmex BTC Swap price liquidations cascade in March 2020) the professional arbitrageurs run out of capital and spread diverges greatly. They know that it will eventually converge, but they have no free capital anymore. This is an excellent opportunity for non-professionals to bet manually on the spread convergence and earn a lot of money with moderate risk.

Sometimes, obvious moderate inefficiencies persists for a long time even under normal market conditions. For example, right now, Deribit June BTC futures trade at 10% premium to the spot price (20k vs 19k). You can just buy BTC for your USD, deposit it on Deribit and fully hedge your BTC exposure by selling June futures. Hold your position until the spread converges (it usually happens in a few months) or in the worst case till expiry. This trade gives you 20%-60% annualised return. Your risk is only the counterparty risk which is not non-existent, but I would say << 5% annualised.

Why does this future spread opportunity persist? I don’t really understand. At the first glance, anybody can take advantage of it. There is nothing fancy here. The asset is very well known. The exchange is rather well known and you can get a slightly worse opportunity on other exchanges like Bitmex. However, when I tried explaining it to my friend who has some investment in stock ETFs he didn’t understand. My hypothesis is that there is just not enough capital in the world understanding such opportunity and being allowed to use it legally and not having even better uses. If it is true, the bar for outperforming “buy and hold stock ETF” is really, really low. 
 

Another piece of evidence against EMH: Coal commodity spot and futures prices have been moving up for several months, with coal stock prices naturally following, but today one analyst raised his price targets on several met coal stocks based on higher expected coal prices, and almost every US coal stock rose another 3-10%. (See BTU, CEIX, ARCH, AMR, HCC.) But there was no new private information released or any change in fundamentals compared to yesterday (futures markets are essentially flat). It's just a pure change in valuation.

Even more damningly, CEIX is up 3.5% (was up 5% intraday) even though it was not upgraded by this analyst, due to the fact that it mines thermal coal and the upgrades were based on higher met coal prices.

SPACs on the whole have been underperforming the S&P 500. This might be because of manager's fees, hype motivating Shaq et al to create terrible SPACs, or low-hanging fruit having been picked early on with hype motivating even competent managers to chase opportunity where none exists.

Being able to pick three SPACs that rose 70-200% in 2 months is still some evidence that you're good enough at picking them to have an edge. But they're not risk-free assets. The risk is that they tie up your money for a long time, only to underperform index funds. With a regular stock, you get to keep your money for yourself, and know exactly what you're buying. Relinquishing control over your money for an extended period of time, and dealing with the uncertainty, both in terms of SPACs being a novel structure and not knowing what the manager will settle on, ought to come with an additional financial reward relative to purchasing stocks. So it makes some sense to me that you'd expect higher returns over a conventional high-risk stock.

Two securities (symbols to come later as this is still being actively traded) are supposed to give the same dividend stream. The company's official website states that they are meant to be economically equivalent.

I don't know exactly what is happening here, and this seems like a strange situation - but there are cases where different securities with identical rights to dividends have different value because they have different voting rights or similar, with implications for pricing if there are rumors of corporate takeovers, for example. Similarly, the fact that they are intended to be equivalent isn't necessarily a binding requirement, and I imagine that  if large investors decide to preferentially buy one class, they could push for changes.

Technically I bought these at slightly above NAV, and brought their effective prices below NAV by selling November call options against them.

How does that work and what’s the downside of that trade?

Re equivalent asset arbitrage, I heard an economist say this was the case for years with Royal Dutch Shell, which traded on both the London and Amsterdam stock exchanges. Even though the prices should have moved around in lockstep they didn't, maybe influenced by other stocks in the relevant country, so could very profitably be arbitraged. Why this persisted for so long I don't know. It's not like it was an obscure little company. Maybe the inconvenience of FX, monitoring prices and trading shares in other countries? (I don't know whether it disappeared with the Internet or persisted more recently.)

My new go-to example of the EMH being false is the merger arbitrage opportunity between TLRY and APHA from earlier this year: https://marketrealist.com/p/tilray-tlry-aphria-apha-merger-date-arbitrage/. I was able to take advantage of this. There are often underappreciated gotchas when it comes to things like this that novice investors miss, e.g. shorting costs or risks of early exercise of options.

By the way, I'd be interested in getting in touch with other semi-pro individual traders in order to brainstorm/share such opportunities privately.

My suspicion about the thin-tailed risk here is that either congress or the SEC passes landmark regulation about SPACs (which is potentially plausible) and those stocks go to 0, very quickly, as the initial investors who IPOed the SPAC pull their money out.  See, ICOs (though those were obviously higher risk)