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 -- discounts are available after registering.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Following myrrdin's suggestion, I'm reposting my questions from the main commodities forum in this thread.
I'm just getting started with futures, don't even have a futures trading account set up yet. I've been looking at commodity ETNs that are based on futures though.
With a very simple strategy I've generated stupendous returns in backtests over the past 5-10 years (depending on how long the ETNs exist). The strategy simply enters/exits ETNs throughout the year depending on what seem to be seasonal patterns currently, as far as I can discern them. This gives me: long platinum in Jan and Feb; long WT crude (or soybeans) in Apr; short silver in May; long sugar in Jun; long palladium in Jul; long heating oil in Aug; short natural gas in Oct, long palladium in Dec. These ETNs have different compositions. Some track a spot price (and hold physical bullion), some the front month futures, and some a basket of different futures. In any case, return with this for 2012-2017: >30% p.a. with no leverage at all. With leveraged ETNs, I can double this at least.
Am I dreaming? Will this turn out a mirage as soon as I start trading this live? And if not, why are these seasonal patterns not being arbitraged away?
Note, I'm aware that ETNs carry unique risks, especially credit risk. Please forget about that. I'm asking a question in principle. Since most of these ETNs track futures I can presumably replicate what they're doing on my own once I have a futures trading account, and won't have to pay their fees.
Can you help answer these questions from other members on NexusFi?
You could certainly test it, and I'm sure you could devise a way to get the 30% annual returns. BUT, could you handle the drawdown involved to get that return, and would your backtest work going forward?
One tricky part to this is that different futures contract months behave differently, because of crop cycles, weather, etc. So you have to figure out which contract month/months match the seasonal tendencies you are looking for.
Well, I *have* devised such a way and it was easy. That's what worries me. How can something like this be so easy? If this were straight equities, it would be impossible.
As for the drawdowns, the worst I've seen was a 25% one in the short silver position in 2009 I think. I didn't mention it in my original post but if I apply an ATR filter to the metals, then I would eliminate all major drawdowns in those. That is, in my test sample, which is just way too short for comfort. And ATR doesn't seem to work very well with the non-metal ETNs.
That's why I'm asking here. Amongst others, I'm hoping that someone will show me some scenarios where this approach would lead to disaster.
For your comparison: MRCI, a well recognized company specializing in seasonal trading for many years, regularly publishes the results of their trade suggestions.
MRCI suggests 15 outright trades per months, and 15 spread trades. In their figures, it is assumed that you trade one future or one spread, respectively. To be able to do that, and use an acceptable ratio between margin and account size, you need a 6 digit account. Slippage and fees are not considered.
Without studying these figures in detail, and without knowing the details of your process, it looks like your selection of trades, resulting in a profit of more than 30 % per year, is significantly better than the selection of MRCI.
Myrrdin, thanks for pointing out the MRCI service. Looking at their results on their home page, they seem far in excess of 30% p.a. but I guess that's because of leverage. Is this correct? I guess most futures trading happens with leverage.
That 6-digit capital requirement you're talking about - is that because the MRCI trades are with full sized contracts, not minis?
Anyone can easily replicate the results I'm talking about. Here are the ETNs and the months I would buy into them each year. All of them have been available at least since 2012. You can get their monthly performance from the below link. If any of the metals ETNs show a significant negative return in a period, say worse than -5%, see if ATR (14, 2.5) would have helped. This you can do on stockcharts.com, for example. Jan-Feb: PPLT. Mar: Cash. April: USO. May: ZSL. Jun: SGG. Jul: PALL. AUG: UHN. Sep: Cash. Oct: DGAZ (divide by 3 to take out the leverage). Nov: Cash. Dec: PALL.
SMCJB, if you're asking whether I've tested out of sample, no. I'm not sure that would make sense here since this is a purely seasonal strategy, with nothing else added. There is no optimization of parameters, there are no trading rules other than the monthly/seasonal rhythm, no stop losses, nothing. What would out of sample achieve here? The past 5 years are what's most relevant, I guess.
What I'm thinking is that since all data are from post-2008, I'm probably working with the best possible sample period. As my strategy is inflexible it would probably perform poorly in a downturn when demand for industrial metals dives.
Trading: Primarily Energy but also a little Equities, Fixed Income, Metals, U308 and Crypto.
Frequency: Many times daily
Duration: Never
Posts: 5,112 since Dec 2013
Thanks Given: 4,473
Thanks Received: 10,323
If your deciding to buy or sell in every month then that would be 12 parameters right there. You've already looked at all the data, so you probably already have biases now, but what I would do, is look at 2012-2016, and then test what you develop on 2017 and see how you did.
Yes, you are right and I retract. But I can't do that. Here's why. The way I've chosen these particular ETNs is
1) Look at https://commodityseasonality.com/ to find assets where the 5-year seasonal trend hasn't reversed course from the 10- and 20-year trends. E.g., copper has reversed trends in several months, incl Jan and Mar. I don't want to bet on something like that.
2) I take the assets that haven't reversed trends, and see if there is an ETN for that. If yes, I check the past performance, especially the past 5 years. If that's been positive then I have a winner.
In order to verify validity of this approach further I would have to have a way to know what those 5-, 10- and 20-year trends were, say, 5 years ago. And I would need the ETNs to already have existed back then with another 5 years' history prior to that. Well, I don't know how to check the first point, and most of these ETNs didn't exist ten years ago....
So basically if I want to trade this, I'm betting on seasonal trends not suddenly reversing across the board. That's not a hugely risky thing to do, or is it? I mean, OK, it happened with copper, but surely it's not going to happen with eight different commodities, all in the same year. That would be impressively bad luck.
MRCI uses full sized contracts. To have enough margin in the account even after a drawdown, I would assume that you need an account of approx. $200,000.
From 2009 until end of 2016 - this is the time frame you are looking at - they made a profit of approx. $160,000, not including slippage and fees. This is a profit of 10 % per year.
I am sure you can do better than this, by selecting the trades more carefully. You find my way to do this in the thread.
