A cash and carry arbitrage strategy applied to bitcoin futures offers annualized returns of at least 25%. Is this anomaly sustainable?
Many investors shy away from bitcoin because of its perceived high volatility or because of its fundamentals.
However, there is an investment strategy that involves buying bitcoin and achieving double-digit returns without being dependent on the trajectory of bitcoin (*).
The strategy in question is relatively simple:
Step 1: buy bitcoin ("long" position) ;
Step 2: sell bitcoin futures contracts at a predetermined date ("short" position);
Step 3: you have entered a market neutral strategy that can earn you a return of between 25% and 50% annualized, without being affected by short-term fluctuations in bitcoin.
Below, we recall the principles of a "cash and carry" arbitrage strategy. We then try to explain the reasons for such a high return and the factors that could eventually reduce the potential return of this strategy.
Futures contracts curve
As a reminder, there are futures contracts on most asset classes. The curves of these contracts depend on their time frame.
The term "contango" curves refers to the situation where the price of a futures contract to its spot price is lower than the price of that contract at a later maturity. The futures curve is therefore upward. Over time, futures prices converge towards the spot price. Therefore, in the case of a long futures position and a price of the underlying that remains unchanged between the purchase of the contract and the expiration date, the investor will incur losses. This is known as a negative roll.
The opposite of a contango is a backwardation curve. In this case, the price of a futures contract close to its spot price is higher than the price of the contract with a more distant maturity. In the case of a long futures position and a price of the underlying that remains unchanged between the purchase of the contract and the expiration date, the investor will receive gains. This is known as a positive roll.
These curve situations are summarized in the chart below. In both cases, an arbitrage strategy called "cash and carry" can be implemented.
In the case of the "contango", it is a matter of creating a "short" position on the "futures" contract (one is then "short" a negative "roll" which translates into a gain) and a long position on the "spot" (in order to cover one's market exposure).
In a backwardation situation, the strategy is to create a long position in the futures contract (to take the positive roll) and a short position in the spot (to hedge market exposure).
Futures contracts curve
The first bitcoin futures contracts on the CME (Chicago Mercantile Exchange) were launched in the later part of 2017.
In the case of bitcoin, the curve has historically been in a "contango" situation. This is currently the case, as shown in the chart below which represents the curve of "futures" contracts on different exchanges.
Bitcoin Futures Curve (source: www.highcharts.com)
The "contango" in question (and therefore the steepening of the curve) is particularly marked, which allows for the implementation of a very lucrative arbitrage strategy. Indeed, as shown in the chart below, the premiums collected via a "short" position on futures contracts can generate annualized returns ranging from 25% to 50% depending on the trade and the maturity...
Note that the returns offered by cash and carry strategies on fiat currencies are far lower than those on bitcoin. With the exception of returns paid on some troubled emerging currencies (e.g. the Turkish lira), it is very rare to generate annualized returns above 5%.
This windfall has not escaped the attention of many hedge funds and prop desks. The Glassnode website shows that open interest in Bitcoin futures is exploding, reaching an all-time high of $23.1 billion late last week. Cash-and-carry arbitrage seems to be behind this craze.
How to explain these extraordinary returns?
Speculation on the rise of bitcoin is often cited as an explanation. But the fact that the curve remains very steep even during bitcoin's strong correction phases seems to negate this reasoning. So we have to look elsewhere for an explanation.
Generally speaking, a futures curve is a function of 3 elements:
1. The spot price
2. The interest rates minus the dividends or coupons;
3. Transportation and storage costs (the cost of carry).
Most bitcoin futures contracts are settled in cash, so the carrying costs can be ignored. Since there is no dividend or coupon, only the interest rate should be taken into account.
The price of bitcoin futures is therefore the cash price plus the cost of borrowing bitcoin over the period.
This cost of borrowing is estimated to be 2.5 - 3.0% per year. The fact that the realized "carry" is much higher than this rate shows that investors must be "compensated" for other risks. Why aren't these generous returns being arbitraged? What are the risks to consider?
The first risk that comes to mind is that of counterparty risk (and cash repatriation) when dealing with offshore and unregulated exchanges such as Binance.
However, this theory does not hold true for a regulated exchange like the CME. And yet, even in Chicago, these returns are very large.
It is then a matter of looking at some of the specifics of the contract traded on the CME, for example. As mentioned before, the contracts are settled in cash and more precisely based on the BRR (Bitcoin Reference Rate) index, which is calculated on the basis of an average of the prices collected on different exchanges in the hour before the close (4pm).
The extreme volatility of bitcoin results in a very high tracking error between spot prices and this reference index. On an annualized basis, this tracking error amounts to 10%. In other words, the high volatility of Bitcoin creates a risk for investors who implement "cash and carry" arbitrage, because their hedge position (long Bitcoin cash position) can see its value vary significantly from the benchmark.
This brings us to an even simpler explanation for the very high risk premium. For many institutional investors, accessing spot positions in bitcoin remains a complicated exercise. Since there is no ETF for Bitcoin, many investors have taken refuge in the Grayscale Bitcoin Trust, which has a market capitalization of close to $40 billion. Problem: the tracking error between the Trust and Bitcoin is 50% on an annualized basis. So it is not an adequate hedging tool.
Which factors could lower these returns?
According to JP Morgan, the launch of bitcoin ETFs in the US could change the game. Cheaper than the trust and with a much lower tracking error, these tools would allow a better control of the risks on the long "spot" side. But they would also allow prime brokers to access collateral more easily and thus lower the initial margin required as a counterparty (it is currently 40%). These ETFs could therefore normalize the pricing of bitcoin futures contracts.
Note that there are currently 7 Bitcoin ETFs awaiting approval to be launched on the US market (see table below).
Table: Bitcoin ETFs awaiting approval (source: Eric Balchunas)
Even though the new head of the SEC, Gary Gensler seems to have a constructive approach to crypto-currencies, it remains difficult to say when these ETFs will be launched if at all, approved.
The extraordinary returns paid by cash and carry on bitcoin could therefore continue for some time to come ...
(*) Historical performance is no guarantee of future results. Even with a hedging strategy, investing in bitcoin and cryptocurrencies involves many risks including counterparty risks. This article is based on market observations and does not constitute an investment recommendation.