Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Total newbie here. Did Futures trading decades ago and retired from Aerospace as a Cost & Schedule Analyst, so I'm good with charts and numbers (31 years). Back then, it was all analog with very little computing like now.
I want to string together a Chart History, for example on Soybeans for a year, two, three even and study the patterns. I understand that back adjusting, changes the opening (?) pricing on the prior contract, on the rollover date of the new contract and then it adjusts going back.
Question 1. Does it adjust the chart data for every single increment, going back?
Question 2. I have a gleaning that this affects the volume shown for prior contract and maybe that's what is being looked at?
Question 3. What advantage does back adjusting offer, versus, just leaving the Open Price (?) unadjusted going back.
Thanks!
Can you help answer these questions from other members on NexusFi?
KevinD, I worked for the Skunk Works for 24 yrs and before that, Lockheed Aircraft Services for 7 (C-130 Mod Kings!). I remember NAA or Rockwell - North American, they built the B-1B up here in Palmdale. I witnessed the two years of B-1B Flight Test and it was awesome! But now, just about everything in Aerospace is slipping across all the companies. They're forcing the 'Grey Beards' out in great numbers.
The B-1B was already flying when I started at Rockwell, although I did work on a few projects for it and had to go up to Palmdale a few times. When I started in late 80s, the old timers then would tell me how things were slipping already from the glory daze of the 50s-70s!
Hope my answer with continuous contracts helps. It is confusing at first, for sure.
Kevin has provided some good answers, just giving some more detail here.
Unlike stocks (for example), futures are contracts and have an expiration. At expiration, the contract will settle at or near the same price as the underlying. For ES, that would be a basket of stocks weighted to equal the SPX index value; for a commodity like gold, it would be the spot price of gold; and so on. The idea is that the contract obligates the seller to deliver the underlying (in the case of physically settled contracts like oil, cotton, etc.) and obligates the buyer to purchase it. For non-physical contracts like ES, which are cash-settled, there is no delivery of course.
The price of a futures contract depends not only on the underlying, but a number of other factors. For equity index futures like ES, that would include interest rates, dividend of the underlying stocks, and days to expiration (see here). For crude oil, that includes the cost to store of the oil until expiration, for example. Generally speaking, if you are storing oil and selling futures on it, your are going to demand a premium for that cost to carry, hence, the further out you go on the crude oil futures curve, the higher the price. This is called contango, and is normal. For ES, it's the opposite, and in a backwardated shape (lower prices in the future, see here). This is because interest rates are so low, and dividends are sufficiently high, that if you buy the basket of SPX-weighted stocks today, you are not foregoing much interest that you would otherwise have accrued, and you are getting dividends high enough to outweigh that, so there is no "cost to carry," there is actually a "benefit to carry," you might say. See indexarb for an up-to-date calculation of these values.
So, this fair value differential is high to begin with, and decreases as time goes on, as the cost/benefit to carry is a function of time. Your cost to store oil is 10 times greater to store it 200 days than 20 days, assuming consistent pricing. The dividends you receive in one quarter of the year to hold SPX stocks is one quarter of what you would receive if you held them a full year. And so on. So, that fair value decreases with time, and at expiration, should be effectively zero, and arbitrage constantly keeps this ever-increasing fair value in line (otherwise, you could buy ES below fair value, sell the stocks at a higher value, and achieve risk-free arbitrage).
Let's do a calculation to make it more concrete.
Yesterday ES settled at 4597. SPX settled at 4605.38. Imagine you bought one ES contract, which has a notional value of $229,850 (50*4597). Imagine someone else paid a notional equivalent on SPX of $230,260 (4605.38 * 50) to buy a cap-weighted basket of all 500 SPX stocks at the same time of day, when SPX was at 4605.38. So, you paid $419 less than your friend. You take $419 and buy risk-free treasury investments. You can get an annualized 0.35% on your money, for 48 days, or: $419 * 0.0035 * 48 / 360, which gives you a whopping 20 cents. Your friend, on the other hand, gets an average dividend of 1.4% on his entire investment, for the 48 days until expiration, or $230260 * 0.014 * 48 / 360 = $429. So, as you can see, you earn no real interest, and while your friend paid $419 more, he got $429 back. Chalk up that $10 difference to my dividend yield accuracy or rounding, and you can see that on December 17, we'll have the same amount of cash. Your friend was compensated for paying more upfront by receiving the dividend. Or, to put it another way, you were compensated upfront by a lower price, as you will not own any stocks to receive a dividend on. If I've made any errors here, I apologize, but hopefully you get the idea.
This is called "convergence," where the futures contract price converges toward the spot/underlying price.
Whew. With this background, you can see that, for the ES for example, over a 3 month span, you simply can't overlay ES and SPX and expect them to match. ES will drift over those 3 months by about 12 points.
When a new contract begins trading with volume (say, the March '22 contract which will do so around December 10), it will, as explained above, be at about a 12 point differential with the previous contract, which will be very close to the spot price at that point. To "stitch together" the old and new contracts, the old one is adjusted so that its close lines up with the close of the new contract on the rollover day. This fixes the "gap" that would otherwise occur. For example, trading December levels one day and March levels the next day, they'll be off a whopping 12 points. So, the old is reduced (in the case of ES) or increased (crude oil, for example) so that the old lines up with the new.
However, when you increase or decrease the most recent contract that is about to expire, then the one prior to *that* will not line up. So, you have to increase/decrease *that* one by the same amount, and presumably it has already been back-adjusted itself. So, what you have, if you do this for many contracts, is a series of adjustments that can cause the chart to show prices that never actually traded, the further back you go.
What's the lesson here? Well, back-adjusting fixes the short term "gap" but adds increasing inaccuracies the further back you go. Some people back-adjust contracts for years, and it just doesn't make sense to do so, IMO. My general rule of thumb is that I don't care about futures prices further back than a year or so. At that point, you need to look at the index or spot price, and if you don't have a good one (aka crude), then you don't really have a good reference.
On top of this, note that the index itself (SPX for example) doesn't take into account dividends. If you compare charts of the SPX total return, which includes dividends, and the SPX spot index, you will see that the march 2020 low to right now is about 4% higher for the TR index, due to dividends. So, even SPX itself is subject to interpretation. If you hold the components in the index, and SPX is 1% lower than this time last year, you're actually up money, since you have received dividends.
So, if you're a short term trader, don't sweat it. Back-adjust, and treat any levels seen prior to 3-6 months with caution. Anything longer than a year or so, treat with extreme caution, and realize that you won't be able to hit a level from 3 years ago to the point anyway.
To add to what @josh says, I would NEVER rely on the actual historical (more than say 6 months ago) prices produced by continuous back adjusted contracts. If you did, you'd see some markets (like Soybeans) actually go negative, which obviously never happened.
But, if you want to backtest for 10 years, then you DO want to test with continuous back adjusted contracts, since if you use a continuous unadjusted contract, you will get incorrect backtest results due to rollover gaps.
Another issue with continuous back adjusted is that using percentage price calculations (such as close divided by close 10 bars ago) may cause some erroneous results also.
All things considered, I ALWAYS backtest with continuous back adjusted contracts. I avoid strategy rules with actual price levels, and I try to mostly avoid percentage price calculations.
Thanks so much for the excellent analysis (and Kevin too). I can see where the adjustment works different ways, for different indexes or futures (ES vs Oil). I also agree with both yours and Kevin's statements regarding how far back you can go, with that adjustment. My take on it, was for example an Inflationary adjustment on the Supply and Demand side of the most recent contracts, that would not be there, like for example, two years ago (but it's 'transitory' right?).
I appreciate the great explanation on this and will take it into my consideration. What I want to do, is a behavioral analysis on individual months and see if there's a commonality between all the Jan Soybean contracts for example, as far as trend formations, volume, etc. Even if it doesn't work, at least I'll understand the action, each month has.
Thanks! Somebody had to dream of this, already. More tools in the tool box! Started re-reading the book I got 25 years ago, and there's still great info in there! Title: Technical Analysis of the Futures Market. By: John J. Murphy.
).