Does anyone know their definition of "fair futures value", i.e. which interest rate and dividend assumption they use? https://www.interactivebrokers.com/.../priceriskanalytics/index_arbitrage_meter.htm
You should ask IB how they calculated that "fair futures value" which is really what this meter pivots on.
If I had to guess, it's OIS for rates and they use announced divs for the companies. It's the same methodology as Bloombergs fair model - it's OK most of the time but tends to miss stuff like moved dates or dividend changes. PS. Unless you can trade EFP, it's hard to take advantage of these dislocations, FWIW.
... Won't the futures price converge with the index at the time of settlement? The index arbitrage meter would show my diminished return or excess return from now to expiration, or not? If I hold the future and invest the cash balance in the money market at rates equal to OIS.
It will converge, barring changes in dividends or funding rates (delta-one desks are usually funded at OIS plus a spread so some of the fluctuations is due to that). However, I fail to see how you plan to make money on that unless you can buy the full basket to replicate the index (that’s what EFP swaps do for you). The arbitrage aspect of equity futures basis is no different from that in say WTI, exept you don’t have to rent a tanker.
I didn't want to make money on futures/index arbitrage; my original question was just regarding IB's tool. However, now that you mention it, if I had for example a non-leveraged account, I could buy the futures plus invest the cash in the money market whenever the futures trade below fair value, and go 100% into the corresponding index ETF whenever the futures trade above fair value. ETFs on major indexes have near 0% expenses. Thoughts? I don't have access to institutional derivatives like swaps. That assumes that the index futures trade below fair value at some time. I don't have data on futures vs. fair value, but I know that almost all equity index futures rolls trade "rich". Maybe index futures don't trade below fair value. Unfortunately IB's tool doesn't currently work for me, so I cannot verify.
You can verify it “the hard way” - load all of the SPX stocks, model the divs and see if the numbers make sense. Actual index arb is hard to do, is very sensitive to funding and, most importantly, is relatively thin. If I were looking for alpha in that space and was ok with capacity constraints, there are frequently opportunities in rolls (for a variety of reasons). It’s especially true if you look at less liquid equity futures or at rolls further away (eg Dec roll in spooz right now).
@Some Lazy Element - Thank you. You seem to be experienced with the valuation and roll aspects of futures, something that I like to study more. I have found little information (about funding rates, deviations from fair value, calendar roll, etc., not the basics) on this forum, as almost all participants seem to be more into technical analysis. For example, I don't really understand how the listed (native) calendar spread order book works - my broker said it's basically hidden. Before we deviate too far from the subject in this topic, do you have any recommendation (books, papers) to read, or is it just learning by doing? I'm not interest in books about the basics of derivatives.
You're welcome. It's not rocket science, most of these things are fairly straightforward and you pick them up by osmosis if you work in the industry. I have to caveat that I've never worked on an index arb desk and only know these things from dabbling in it while on the buyside - so I might be off on specific details. Various broker-dealers have introductory papers on these things as part of their research portals (JPM and BarCap are particularly). If you have anything in particular that you want to me to look up, lemme know.
Thanks, Some Lazy Element. One question that I've not been able to find an answer is how the "complex order book" of listed (i.e. native) futures roll calendar spreads works. I haven't been able to find a description on the futures exchanges' web sites. I am observing that IB (for example) only quotes the sum of the spreads of the individual legs, but I am told by the brokers that I can get better fills, i.e. the existing bid and ask of the listed (native) spread product itself is sometimes more narrow, but the bid and ask of the (native) spread itself is "hidden". I'm not sure if this is all that there is to it, and why the order book would be hidden. I am also told that "complex order book" means that the native spread product "can" be matched against the regular order books of the front and back month future contracts, but I'm unable to find an accurate description on the "if and how", priorities, etc.