On the face of it this sort of pure arbitrage should be impossible, but it might make sense if for example you can do a trade that others can't because of constraints. With that in mind, let's run the numbers.
VXX is invested in the first two months of VIX futures, with some kind of gradual rolling schedule, for example it's currently 13% in September and 87% in October. On the face of it we can perfectly hedge the underlying index, although not the ETN itself, by replicating that futures position. That also increases the minimum trade size for this. You can hedge 1000 shares of VXX with a single VIX future, but if you're going to roll in say 10% increments you need to buy 10,000 shares of VXX so this is quite a chunky trade ($200k notional).
(If you do this with a single contract, you will have some basis risk which could be quite large as the front of the VIX is quite spicey)
Total 'investor fee' on the ETN is 0.89%. I'm not familiar with the US system for quoting costs, so that might for example exclude transaction costs on the future but they are unlikely to be material. Similarly trading costs on the futures hedge will knock your returns down a bit, but not very much.
There were a regulatory issues with Barclays ETNs that messed up the premium relationship as the article says; 6% still feels large for an ETN that is easy to hedge.
So if:
- the premium doesn't narrow
- and our shares are lent out
We expect to pay 0.89% and earn perhaps half of 34%, 17% for a net of ~16%, maybe knocking off 1% or so for costs.
We also have to put up margin for the VIX trade, plus have money to cover potential losses in the future which were offset by unrealised gains in the ETN. That increase the capital you need to employ and thus reduces the rate of return.Looks like the margin on the future is $16k per contract, so that implies you'd need another $160k just for the margin on 10 contracts; I'd probably want another $40k in hand, so the total capital required is double what I said above ($400k) and the return is more like 7%. I don't know if IB, for example, would partially net the hedge off in a margin account, in which case you could probably do this trade with a lot less than $400k. That would increase the potential returns, but also the potential risks.
The key risks on this trade are:
- The premium vanishes (which would knock the gross return down from 16% to 10%, and the return on capital down to 5%) or worse becomes a discount (if the discount went to 6% the the gross return would drop to 4%, and on capital just 2%).
- The shares aren't lent out. If the shares aren't lent out you're basically on the hook to pay the ETN fee of 0.89% plus some costs, so maybe the downside isn't very big at around 1%.
- Both, which seems unlikely. But if the shares weren't lent out you'd lose 12% if the premium also went to a 6% discount, plus 1%, which is 13% or 6.5% on capital
- In the event you can't get a hedged margin benefit; a very large rise in the VIX requiring you to post more capital to cover your futures losses, or risk being knocked out of the trade.
So this looks like a reasonable trade (upside of 7%, versus downside of around 1%, maybe 7% in a black swan situation, with the possibility of leverage to juice those numbers up), but not a slam dunk. Returning to the original question, why should such a large borrowing fee exist? Maybe there are a lot of people who want to do the opposite of this trade (so short the ETN and long the future); perhaps because they think the premium will vanish. But the sort of people who do this kind of trade in size wouldn't be interested at such a high borrowing fee which completely wipes out the 6% upside if the premium goes to zero. In fact, it's likely the borrowing fee 'on the street' for institutional counterparties is lower; IB is offering the borrowing service at a 17% - 34% bid/offer, suggesting the mid is around 26% and the institutional numbers are probably more like 23 - 29%.
More likely, there are people who want to short VIX but can't trade the futures. So instead they short the ETN. The fact this is extremely expensive doesn't bother them, as these sorts of people are probably meme stock muppets and don't actually understand the finer detail. Plus they're probably doing the trade on a short term basis, so the borrowing fee won't bother them.
Anyway, TLDR: Would I do this trade? Probably not.
GAT
