If you ever had a look at CME’s Bitcoin futures, you might have noticed that there are multiple contracts - one for each month, and that they have different prices. Systematically, the futures contracts for months that are further away, are more expensive than those with closer expiration, and also cheaper than spot. When this happens for a commodity like oil, we’re talking about contango.

This is usually due to the fact that there’s a risk premium on the uncertainty of the future, and it costs a little bit to be sure of the price that you’ll pay for something three months from now. But for futures on financial products, it’s very rare. Currently, the premium for Bitcoin is a whopping 2% per month, and it’s a red warning sign.

Term structure of Bitcoin futures

The reason contango for commodities like oil is normal, is that you can’t arbitrage it. The way to arbitrage contango is to buy spot, short the future, wait till expiry, and collect the premium. But you can’t simply “wait” when you buy thousands of barrels of oil - you have to store those barrels somewhere, and it will cost you an arm and a leg. The storage costs will eat up your contango premium, plus some.

There were occurrences in the past where traders were able to arbitrage this premium, but the contango needed to be absolutely off-the-charts (“super-contango”). For instance, after the 2008 financial panic, the spread between spot and futures in the oil market was so big, traders would literally rent tankers and use them as storage facilities for oil, waiting for their futures leg to mature.

The reason contango for financial products isn’t normal, is that you can arbitrage it very easily, as you don’t have storage costs. You buy Bitcoin spot, sell an April future for 4% more than spot, and simply wait two months.

At expiry, you’ll sell your Bitcoin at a certain price, while the future will settle at the exact same price, and you’ll collect the premium. Your only cost is the carry, i.e. the interest you have forfeited on the cash you used to buy Bitcoin, but this is virtually zero today.

Why the premium, then? Why don’t hedge funds with hundreds of billions of dollars to spare don’t arbitrage the Bitcoin contango down to zero?

Credit risk?

One quick answer would be that you aren’t guaranteed to close your trade because of counterparty risk. This is what happened when Lehman went bust - all the hedge funds that had OTC derivatives positions with Lehman as one leg of a trade, suddenly saw the “Lehman leg” disappear, while the other leg was still there. Lehman going bust single handedly destroyed the convertible bonds market, as hedge funds who had bought convertible bonds and hedged their risks (credit, volatility, equity risks - you name it) with Lehman, suddenly were out of hedges, and had to sell their cash position (their convertible bonds) at literally any price.

But there is no counterparty risk here. You can’t lose your Bitcoins - you just keep them on chain, and no exchange going bust will affect your position. CME can’t go bust - it has margin calls, initial margin deposits, has a guarantee fund, and is almost explicitly guaranteed by the Fed because of its critical importance to financial markets.

So, again, why is nobody picking those hundred dollar bills lying on the street?

Systemic risk

There is one risk that you can’t edge - that of the world collapsing overnight, when there’s nobody left to pay you even if you win your bet. Imagine shorting USDT on Binance. What happens if you’re right and USDT goes to zero? Do you really think Binance will still be around to pay you?

You can always say, “bah, I’ll see it coming, the moment shit starts to hit the fan, I’ll close my positions”. But what if you don’t have the time to do so? What if a black swan event happens overnight?

This is the risk with Bitcoin. A whim from regulators, like passing the STABLE act, and all USDT exchanges go bust. What will happen to Bitcoin futures then? CME uses the “Bitcoin Reference Rate” (BRR) to calculate the spot for its Bitcoin futures, which is a weighted average of the spots taken from multiple crypto exchanges, the list of which changes over time. What happens if the Bitcoin ecosystem collapses, the prices of Bitcoin on half of those exchanges go to 100,000 on paper because those exchanges become insolvent and people rush for the exits by buying Bitcoin to try and send it to their own wallets (as happened with MtGox)? Futures skyrocket because of a technicality, CME is slow to react, and you get wiped out on your “short futures” leg - while you can’t really sell the “long Bitcoin” leg either, because every exchange is on fire and nobody processes redemption requests.

Implied probability of default

This risk is rewarded handsomely - a big fat 2% per month. With interest rates at zero, we can borrow a formula from the Credit Default Swaps market to calculate the implied probability of default, which we’ll simplify to the extreme:

p = S/(1-R)

Where p is the probability of default, S the annual “swap rate” that compensates for that probability, and R the recovery rate, i.e. how much you can expect to recoup from your investment if a default occurs.

In our case, S = 24% (12 months in a year, at 2% for each month), and R ≈ 50% based on what happened after MtGox collapsed. This means that we assume you’ll be able to cash out in the event of an apocalypse with an average 50% loss. Remember, this is assuming that you are hedged!

This gives us an implied probability of default p ≈ 50% over the next year. Again, that’s the probability of the whole Bitcoin ecosystem collapsing overnight, not giving you the time to close your positions, cutting you off from exchanges, and where CME futures no longer trade.

This is the risk as it’s being priced by the market, on relatively liquid financial instruments. So the next time someone tries to explain to you that “Bitcoin is a store of value”, when the chances of it getting utterly destroyed overnight before the year is over are around 50/50, try not to shoot coffee through your nose laughing.

Leave a comment

Your email address will not be published. Required fields are marked *