Selling the Froth: A Simple Hedged Forex Strategy for Bitcoin-Denominated Returns

Why would the Bitcoin markets permit a simple forex investment strategy to remain persistently profitable even in the face of quite outlandish changes in the value of Bitcoin versus the dollar? There are at least two possibilities: either the market is inefficient, or the market doesn’t think very highly of the futures exchange involved.

The Simple Theory: Selling Froth

The strategy is simple: when Bitcoin futures markets price in a rate of growth in BTC versus the dollar which significantly exceeds an investor’s cost of capital for borrowing dollars during the same period, the investor buys bitcoins with borrowed dollars and takes a counterbalancing short position in the relatively high-priced futures. Provided the position is maintained until expiry, the range of movements in BTC versus USD over which this strategy is profitable depends on just two factors: the cost of capital, and the initial premium built into the futures which are to be shorted. The primary risk to the strategy is the counterparty risk of the futures exchange, while a secondary risk is that should the position need to be liquidated prior to expiry, the premium built into the futures may have expanded rather than contracted, requiring additional capital to close the position.

(In this context, I am deliberately avoiding the terms contango and backwardation, which are often used incorrectly in Bitcoin circles to describe the premium or discount, respectively, of futures relative to current spot. In normal futures lingo, contango and backwardation refer not to the premium or discount relative to current spot, but relative to the expected price of the commodity at settlement.)

To unpack this a little, the strategy rests on the fact that by the time of settlement, futures and the spot market must converge. Knowing that any premium relative to the spot market must eventually dissipate, the investor shorts futures when that premium becomes relatively large. Naturally, the spot market may in fact move upward — perhaps strongly — during that same period of time, so to balance the exposure created by the short position in futures, the investor adds long exposure in the spot market by buying additional bitcoins. (Of course, the investor needn’t actually borrow any dollars to buy bitcoins: borrowing bitcoins with a fiat-linked repayment agreement will achieve the same end result and will do so more efficiently.) If the spot price does move upward, the investor will need to exchange only part of that borrowed position in more valuable bitcoins in order to repay the fiat-denominated loan. Likewise, the spot price may move downward, and the liability created by having borrowed dollars to buy bitcoins that become less valuable is therefore balanced by a short position in futures: should the spot price fall, the additional bitcoins required to repay the fiat loan will be provided by gains in the futures. A primary benefit of this type of strategy is that it specifically does not require making a directional bet on BTC vs. USD; when the futures market is sufficiently frothy, this strategy can be profitable regardless of whether BTC rises or falls versus the dollar.

Here’s an early and simplistic version of a spreadsheet I’ve sometimes used to evaluate the strategy in the course of managing Bitcoin-denominated hedge funds; the total fund size and capital to deploy are just dummy placeholders, at 100 BTC and 25 BTC respectively, but the pricing and futures expiry data are real examples, taken from Bitstamp and ICBIT while I was preparing this article:

Simple long BTC, short futures strategy

If you click on the image to view the original, it’s easy to see that for an investor with access to fiat capital at 10%, the strategy of buying bitcoins and shorting futures is profitable in BTC terms all across the range from a 90% fall in BTC vs. USD to a 400% gain in the value of the cryptocurrency. This simple version of the spreadsheet shows returns relative to twice the amount of deployed capital — in this case, borrowing 25 BTC worth of fiat means the returns are shown relative to 50 BTC — as well as relative to an overall fund size which might be larger than the capital deployed for the strategy. The spreadsheet shows a return of around 9 BTC (about 19%) if the exchange rate remains stable, a return of 17 BTC (about 34%) if Bitcoin falls by 40%, or a return of nil if Bitcoin should rise by 400%.

Obviously a speculator or gambler who believes they can predict the future could instead simply take unhedged positions, making entirely directional bets, and they could achieve much greater returns by doing so. The point of this strategy, however, is that one does not have to be right about the direction of future exchange rate movements in order to profit, and one can profit handsomely even while the exchange rate remains steady.

Notably, this type of strategy was viable for many months until the administrators of the ICBIT exchange made it their goal to force futures prices down relative to spot toward the end of 2013; relatively recently, however, the strategy has again popped up as potentially worthwhile via the September 2014 futures contract, which is the one shown in the spreadsheet screenshot. It’s especially notable that on some occasions, it has actually been possible to execute this strategy using capital borrowed at the going market rates on Bitfinex, with no additional external source of capital needed.

Bitcoin Forex Strategies in Practice

In real investing, merely plugging a few numbers into such a simple spreadsheet could be a recipe for disaster. More sophisticated approaches would assign a risk premium to each of several factors which could negatively impact the strategy, such as the counterparty risk of the futures exchange itself: should the futures market fold, not only would the investor’s margin capital be lost, but gains in futures would no longer be available to help repay a fiat margin loan in the case of falling BTC values. In addition, if there is any chance the position will need to be liquidated prior to settlement, an investor would want to have a good handle on the extent to which the premium reflected in the futures market might actually expand rather than contract. This unpredictability in futures pricing is very much akin to changes in implied volatility in options pricing: in the short term, it’s entirely possible, for example, for rising implied volatility to drive the price of a written call higher even while the price of its underlying equity is falling or remaining steady.

(In fact, careful management of a strategy like this — with a view to adding or subtracting small amounts of delta in response to changes in the exchange rate — can potentially capture even more gains by playing off exactly these types of changes in the premium. Generally speaking, futures premiums tend to expand and contract differently in response to movements in the exchange rate up or down.)

Finally, a strategy like this one would rarely occur in a vacuum; to the extent that an investor is looking to execute an overall strategy which involves this one as a component part, the job of analyzing potential outcomes and evaluating risk becomes all that much more involved.

What Does the Viability of Such a Strategy Tell Us?

The fact that this type of strategy is ever viable at all seems to me to suggest at least one of two possible explanations: either the futures market is relatively inefficient and dominated by unhedged directional bets taken by gamblers and speculators, or the aggregate view of the Bitcoin capital markets is that the counterparty risk created by exchanges like ICBIT is too high to support strategies that require maintaining a position thru to settlement in order to ensure profitability.

Looking at the first of those two explanations, it’s a general rule of thumb that activity in a given derivatives market which creates opportunities to profit by way of activity in the underlying spot market will attract market participants to exploit those opportunities (and vice versa). If someone is willing to pay a great deal more via futures than in the spot market, then one would expect other market participants to step right up and meet that demand, pocketing the difference relative to spot. However, the fact that the market frequently permits the two to drift so far apart that one can lock in Bitcoin-denominated gains that survive even a 400% rise or 90% fall in BTC vs. USD suggests an imbalance between unhedged long bets in the futures market and spot market participants willing to exploit and profit from the futures market bullishness.

And looking at the second of those two explanations, there are few better reasons for avoiding strategies which appear on paper to be high probability winners than the risk that some other factor which is not part of one’s model is going to cause the whole thing to come crashing down. In other words, if investors are not taking this opportunity to generate a modest return which is relatively safe against changes in BTC vs. USD, it suggests that they believe the return is in fact not safe against some other factor, the most obvious of which is the risk represented by the exchange itself. Anecdotally, my experience of attempting to adjust positions in response to incessant meddling and rule-tweaking by the administrators of ICBIT certainly undermines my own confidence in the exchange’s reliability. (If the exchange can roll out four major changes with significant market impact in the space of three months, how many might it introduce over the next seven?)

Probably neither of these two alternatives reflects particularly well on ICBIT as a reliable tool for investors.

A Personal Note: Why Publish a Strategy That Has Been Working?

It’s well established in the literature that publishing an investment strategy which has been working by exploiting an apparent market inefficiency relatively quickly results in that strategy no longer working, as other market participants move in to eliminate the inefficiency. So if this strategy has been working, one might wonder, why would I discuss it publicly?

For a start, this particular strategy isn’t rocket science, and nor, as far as I can tell, is it unknown. Many traders and speculators familiar with ICBIT routinely talk about taking opposite positions in the futures and spot markets — and such hedging is, after all, largely what futures markets are for — sometimes erroneously referring to the activity as “arbitrage” (which it plainly is not), so it seems highly unlikely that nobody would have noticed this particular strategy. I think it is far more likely that this strategy is actually no big secret, that it is “old hat” even, and that the strategy’s apparent viability is due less to inefficiency than to a relatively negative aggregate market view of the trustworthiness of ICBIT.

In addition, while I ordinarily would not publish a discussion of any strategy which might be directly relevant to my own Bitcoin hedge funds, regardless of what I might think of market efficiency, in this case, I see little danger to fund participants from doing so.

Finally, on a personal note, I find myself sufficiently disappointed in the experience of managing Bitcoin funds that I may well elect not to continue with the activity in the short term. For someone attempting to operate at the conscientious end of the spectrum, amid a seemingly endless supply of hucksters running outright scams, the job of looking after other people’s Bitcoin-denominated capital is quite a lot more time consuming and stressful — and quite a lot less fun — than I had imagined it might be. In my experience, managing funds of up to a couple thousand bitcoins simply does not offer sufficient financial returns to make up for the stress of swimming through all that muck while safeguarding other people’s capital from loss.

It doesn’t help that the Bitcoin capital markets currently appear to be significantly less mature and reliable than even the worst of third world markets in fiat finance. The Bitcoin markets are currently overrun with piles of garbage “investments” promising — or at least very strongly suggesting — returns which they will in all probability never generate. And creating additional irritation from a fund management perspective, returns in the BTC space are still very strongly influenced not just by investment approach or competence but by quasi-random factors — ranging from abrupt exchange closures, capital controls or rule changes to blatant fraud and criminal behavior — that have utterly nothing to do with investment strategy or the normal functioning of a reliable financial market. If someone is attracted by the mathematical structure of financial markets or by the opportunity to create and execute well designed investment strategies, and especially if they have successfully cut their teeth out there in the real world of fiat-denominated investing, the third world qualities of Bitcoin capital markets hold limited appeal. Bitcoin itself certainly continues to appeal, and I count myself among the many entrepreneurs and investors who are very excited by the opportunities it presents, but those opportunities are another kettle of fish altogether: Bitcoin itself might have a great future and be very worthy of investment, but that doesn’t mean that most so-called “investments” in what pass for Bitcoin capital markets are anything other than utter junk.

I know for a fact that I’m not alone in getting more than a little bored with what passes for capital markets in the Bitcoin space. So, if publishing a snippet of something which I find interesting about the dynamics of futures versus spot markets generates a bit of interest from someone else out there, then that’s good enough for me.

• asset pricing
• Bitcoin