2

If you look at, e.g. the best offers at the 420 BTCUSD strike price (this was as-of 4:15 pm EST on November 14, 2013), or at the "How It Works" page for Future Block, it seems like there is a violation of the No Arbitrage Rule.

For example, from the "How It Works" example. Suppose I sell one BTC's worth of "up" contract (I get 1 BTC) at 80% payout (I pay back 1.8 BTC if price is above strike at expiry).

Next, I sell one BTC's worth of "down" contract at the same strike price (I receive another 1 BTC), at 50% payout (I pay back 1.5 BTC if price is below strike at expiry).

If the price is above the strike, the "up" contract clears and I received 2 BTC but pay back 1.8 BTC (netting a gain of 0.2 BTC for me). If the price is below, the "down" contract clears and I receive 2 BTC minus the 1.5 I pay back (netting a gain of 0.5 BTC for me).

Obviously, these example offers may not be close enough to the best offer to attract buyers, but apart from that I can't see any specifications for transaction costs to the seller, fees, etc. Even if there were fees, wouldn't they need to adjust according to the payout rates a seller could command selling on both sides of a strike price?

I think I am probably missing an obvious detail; any help correcting my mistake would be great!

Added:

To clarify, the puzzle here is that at about 4:15 yesterday, the best offer for up and down contracts roughly matched the example I give above, using numbers from the "How It Works" page to illustrate a Dutch book offer selling on both sides of the strike price. That this can happen is not a mystery per se, but that the live best offers reflected that it was a viable live strategy even when there was resonable volume in the market is a puzzle. Either there was simply no volume at all on one side of the strike price (which should have immediately led to better offers on that side) or else there was actually not much participation and the volume data was incorrect / misleading, or both. (Assuming that early adopters of BTC options can't be that irrational... after all, they'd have to know a lot about bitcoin and that particular options platform even to participate. So it seems implausible that buyers were so irrational as to be buying at such prices on both sides of a strike price).

1 Answers1

2

It is a well-known observation that Bitcoin exchanges are not efficient (in the academic sense of well-approximating what you call the No Arbitrage Rule). The reason is simply that there are real world barriers that academics usually assume not to be present. If you do consider the risk associated with depositing funds at a Bitcoin exchange, there is an obvious reason why someone would demand a high return for taking such a risk. If you look e.g. at the difference in exchange rates between big exchanges (try Mtgox, Bitstamp, BTCE) you'll see that a difference as large as 10% can persist on the timescale of a year.

The risks include having the exchange hacked and leaving it unable to return your deposit (have a look at Bitfloor) or having its bank accounts frozen during legal investigations and hence unable to return your deposits during the first 6 month or so (check bitcoin-24.com). And then there's the combination of it all with (presumably) enough reserves not to shut down the operation right away (e.g. mtgox). And that's not even considering the risk that the operator is dishonest and gets tempted to defraud customers...

On top of all this, consider that even moving Bitcoins around is not instant. You run the additional risk that during the transfer time the situation changes (e.g. a crash in Bitcoin exchange rate). And the transfer time between bitcoin trading sites can be surprisingly long, since some legislation (e.g. anti money laundering) typically seems to compel them to insist on verifying your identity. This process may, depending on the site, occur in a somewhat surprising fashion, so you never know if they will instantly, or at least within the first days, allow you to continue accessing your bitcoins (or real money) that you gave them.

  • I'm not sure that I buy that any of these points (which have perfectly suitable analogs in many other low-liquidity, risky markets) matter unless bitcoin options buyers are *especially* ignorant of properties of bitcoin. If the buyers are remotely rational / knowledgeable about the tech, then something as simple as no arbitrage should be immediate. Plus, the volume numbers at FutureBlock seemed reasonably high. –  Nov 15 '13 at 14:21
  • If there is real risk, it does not require ignorance. And if there wasn't, then I suppose this should be a decent suggestion: How about you take advantage of the arbitrage opportunity, and tell us about your experience once you've been at it long enough to have statistic evidence about what risks you took to get these profits? –  Nov 15 '13 at 15:59
  • I think you're misunderstanding. There would be no risk to the seller. That's the problem. I could offer to sell on both sides such that my offer was the best offer for both the up and down contracts, and as long as there is one buyer for each, I am guaranteed to make money. Yes, there is liquidity risk, as in many other markets, which ought to be "priced-in" if the buyers are not especially irrational. –  Nov 15 '13 at 16:22
  • I see. You have zero risk that FutureBlock runs away with your Bitcoins, breaks its promises, or otherwise defrauds you. And you have zero risk that whilst there's no fraud there is some misunderstanding that ends up costly or otherwise tedious to clear up. You also have zero risk that, say because of a large change in Bitcoin price, your indeed gained Bitcoins will not be worth anything anymore. Lucky you! If I were to trade on the opportunity you found, I'd feel like I was taking on a significant risk and would want to insist on a large return for that. –  Nov 15 '13 at 16:28
  • Eye roll. None of which effects no arbitrage. Whatever such risks are, they don't imply that buyers would tolerate best offers *on both sides of the strike* that imply a dutch book. You seem to think these kinds of risks don't exist in other markets (but they do) and that apparently people tolerate dutch books magically because there is risk. –  Nov 15 '13 at 16:35
  • There is no magic involved. If there is risk, a rational trader will demand extra profit to compensate him for the risk. Only if arbitrage is free of risks, or so profitable as to more than compensate for the risk (and effort, especially in illiquid markets as this) we can expect to usually find someone willing to take the arbitrage, hence moving the market towards a no-arbitrage condition. –  Nov 15 '13 at 16:48
  • 1
    I understand. I do not believe that exchange insolvency risk explains this instance and am seeking other explanations. Or, legitimate evidence that it is exchange insolvency risk preventing additional sellers from taking advantage. –  Nov 15 '13 at 16:52
  • Here's the historic evidence for the insolvency risk of bitcoin exchanges: At least 2 of the 5 largest exchanges have become insolvent (as in permanently stopped trading and failed to fully pay back user deposits) during the past two years alone. These were bitfloor and bitcoin-24. Seems like a very significant risk to me. –  Nov 15 '13 at 17:26
  • These are options on BTCUSD. –  Nov 15 '13 at 18:06
  • My examples are not, but both the exchanges from my example and the sites offering options share similar risks based on taking custody of customers' bitcoins (or other customers' funds). By the way, the numbers may be even worse for sites offering some kind of derivative (options!) or leveraged trading. Here the famous insolvency from within the past two years is Bitcoinica. –  Nov 15 '13 at 18:22
  • But since the market participants, by definition, had to already believe it was worth the risk to store money in bitcoins, and since the options market is about the exchange rate, this wouldn't matter. By definition, the contributors to volumn at that site are folks already fine with the risk of having money in bitcoin formart. And it is these folks, who already have bitcoins and are buying options, who are allowing the arbitrage opportunity to persist. –  Nov 15 '13 at 18:24
  • I'm sorry if that's not clear, but it doesn't make sense to say that bitcoin exchange insolvency risk would prevent people who are already fully bearing that risk and already expressing willingness to buy options on BTCUSD from exploiting an arbitrage opportunity. Another point is that BTCUSD is partially *a function of* exchange insolvency. So to the extent that someone believes that's a credible risk, they ought to be willing to bet on the exchange rate plummeting due to exchange insolvency, much the same as someone trading currency swaps in extremely illiquid, corrupt frontier markets. –  Nov 15 '13 at 18:25
  • If that were the case, there would not be any arbitrage opportunity. Clearly more funds need to flow into the site for that. Apparently the people who have existing funds on the exchange have invested these in a way that no longer permits arbitrage (maybe they need to use all of them to hedge their exposure to the BTCUSD exchange ratio elsewhere), or they wait because they expect to get an even better arbitrage opportunity in the future. [Or they act slightly irrational, as real world traders sometimes do in all market.] –  Nov 15 '13 at 18:35
  • Exactly! Now you're getting to better explanations! –  Nov 15 '13 at 18:55
  • Sorry, but I don't understand what you mean with "getting to better explanations." As far as I can tell I haven't changed my argument one bit, but only spelled out the details that I initially thought were obvious. Maybe the problem lies with having different ideas about what assumptions lead to the No Arbitrage Rule? –  Nov 15 '13 at 19:05
  • The generic risks that are ever present around bitcoin exchanges, insolvency, etc., can't logically be the explanation for why a bunch of folks already storing money in bitcoin and already providing volume in BITUSD options would fail to exploit an arbitrage opportunity. Your answer and subsequent comments contained only these logically invalid points. But later you realize and offer different points that don't have much to do with insolvency risks and I assess these as being more likely explanations, though still not fully satisfactory. –  Nov 15 '13 at 19:15
  • So this is the misunderstanding: I do not believe there is a logical invalidity. If, as I think we agree, there is inherent risk in these deposits, then any rational actor will decide to keep deposits only if the reward is sufficiently large. That logically implies not taking every (otherwise risk free) trading opportunity, but to withdraw or otherwise withold funds until a sufficiently large opportunity arises. Hence if the textbook zero-risk is replaced with real-world risk, "No Arbitrage" becomes "some [possible persistent] arbitrage." Even, to a lesser extent, in larger real-world markets. –  Nov 15 '13 at 19:23
  • The reward does need to be sufficiently large relative to the *relevant* risks, yes. But, e.g. bitcoin exchange insolvency is not relevant conditioned on a person already, in the face of that risk, obtaining bitcoins and buying/selling options on BTCUSD. To buy bitcoins implies willingness to accept the exchange insolvency or market clearinghouse insolvency or legal risks associated with placing one's money into that state. Given that, the next step, to buy/sell/do nothing, is independent. E[E[A|B]|B] = E[A|B] –  Nov 15 '13 at 19:32
  • There is a difference between the risk of "putting money into bitcoin" (and storing it in your offline paper wallet) and the _additional_ risk of entrusting these bitcoin to an industry with a high insolvency rate. And that later part is the relevant risk because at any time a trader has the choice between withdrawing his bitcoin from the site or leaving them on there for trading (e.g. in your arbitrage). –  Nov 15 '13 at 19:50
  • Fair enough. I guess I assess the incremental risk beyond that incurred just by putting it in the wallet (among people already engaging in BTCUSD options trades) as so small as to be entirely neglible. We may just assess that differently. –  Nov 15 '13 at 20:14