Here’s part two of my interview with Tim Pickering, president of Auspice Capital Advisors, who manages both the Horizons Auspice Managed Futures Index ETF (HMF) and the iShares Broad Commodity Index Fund (CBR). You can read part one here.
Managed futures has traditionally been a hedge fund strategy, used mostly by institutions. Can it really be adapted for the retail ETF investor?
TP: All we’ve done at Auspice is said, this is what we do in our alpha program. It’s been around, it’s got a great history. So what are we are willing to make transparent by putting it in an index? How far are we willing to lift the kimono? Can we do that with an ETF-like price? We started talking about that years ago and people thought we were out of our minds. Nobody has done this before, and we are definitely getting a lot of eyebrow raising—that’s a polite way of putting it.
When it became known that we were publishing these broad commodities and managed futures indexes, I got a call from the CEO of a major company in Canada that has a managed futures program. He said, “Tim, you are about to wreck the managed futures space in Canada,” because of these indices, and because you can get them with a 95-basis-point ETF. And I said, hold on a second. First of all, what managed futures space in Canada? Nobody even knows about this in Canada. It’s the number-one alternative strategy in the world, and nobody uses it in Canada. And second, if you are good at generating alpha for fair price, then you’re fine—the ETF is not going to hurt you. But if you’re just expensive beta, or if you’re overpriced and underperforming, then you’re screwed. And whether I do it or someone else does it, this is the evolution of markets.
The Horizons Auspice Managed Futures ETF (HMF) is an index fund. Does that mean all of this is done quantitatively, without any hunches or intuition from the manager?
TP: Exactly. Hunches and intuition are for amateurs. We follow a three-step process and each of the those steps can be quantified and replicated, and they’re published in the index methodology.
The first is the trend following algorithm, where we look across all of these 21 positions and identify whether the market is going up or down. Then we decide whether to participate in the market, and do we go long or short.
Step two is position sizing and rebalancing, based on volatility. Once we know whether the market is going up or down, the question is how big do you trade, and when do you adjust that position size. Then we normalize the risk across all of those assets: we want to take the same dollar risk in natural gas as we do in five-year notes, and cotton, and all of these other things.
Going forward in time the risk of those assets changes, because the volatility doesn’t stay static. So let’s say we put on natural gas and it starts to trend up, and volatility goes from 30% to 50%. A month later when we look at it, we need to make a decision on that position size: should we adjust it? If it meets the threshold we use, then we rebalance that position back to where it was. The index calls for this on a monthly basis.
Finally, the third step is rolling the futures contracts. We look at the forward curve and we say, is there a more intelligent decision we can make than simply rolling month-to-month? That’s always a decision we have to be dynamic about: where we want to be on the forward curve based on the risk-reward in the market. We have a process and methodology that we follow, and I don’t need to be staring at a screen to do it: we’ve taken our experience and programmed it, so we are consistent about it, and when it comes to executing we can do it systematically so we don’t mess it up.
The iShares Broad Commodities ETF (CBR) uses the same strategy, but instead of trading commodities and financial futures long and short, we just trade the commodities long. If the trend is going up, we go along, and if the trend is going down, we go to cash. Otherwise the three steps are the same.
One of the problems with ETFs like this is that they may not track their indexes well because of fees and transaction costs. Will we be able to measure the tracking error reliably?
TP: Yes, absolutely. Because the step we made at Auspice—why that CEO was ticked off at me—is that we published the index. You can download it from Bloomberg or Reuters and get your numbers daily, and see the holdings and the movement of that index. And HMF should follow that quite closely, keeping in mind that the fee is 95 basis points.
One of the interesting things about what we do is that this is not a very active strategy. We are taking a medium- to long-term trend following approach, so there’s not a ton of trading. And because we position ourselves on the futures forward curve that also eliminates transaction costs: by positioning ourselves properly we don’t have to overtrade. For example, I could trade every month in natural gas, or I can just put myself in January, April, and December. So not only are we not trading a lot, by just being long and short, we are positioning ourselves to eliminate some of that transactional aspect.
Futures are also great because they are so cost-effective: it costs me a couple of bucks to round-turn futures. The thing that can cause deviation is slippage: you go to buy crude at $80 and by the time I execute it, it’s $80.50. That’s a risk, and that’s where we have been hired to do our best job. We have been hired for that execution ability as well: hopefully we are good at that.
The other part of the equation for HMF is this: because of the leveraged ability in futures, I only have to put a certain amount on margin, and the rest is in cash. Most of that money in cash—that 90% to 95%—is earning a return. Not a lot in this environment, but 100, 120, or 130 basis points.
This sounds active enough to me to not be part of my portfolio. “Hunches and intuition are for amateurs” then goes on to say “then we decide whether to participate in the market, and do we go long or short”. Sounds like that decision is based on a hunch to me. My interpretation of Mr. Malkiel’s “A Random Walk…” is that no “trend following algorithm” is going to work consistently enough to be able to time the market.
Managed futures are an alternative strategy, to the usual long stock/bond portfolio. I’m unsure about managed futures, but I am interested in alternative strategies. My problem is the punitive tax I pay on bonds. I am interested in alternative strategies that have bond like properties, but are more tax efficient.
@SterlingF:
I barely understand futures anyway, but am reading this to try and broaden my horizons. If futures could be intellectually boiled down to be a highly volatile but objective component of a Passive Index plan, then in theory I could see that as an ingredient CP’s could use. But I think your analysis of his dodging and weaving are spot on. “Hunches and intuition are for amateurs???”. But that’s his method! Definitely Active Management.
Sterling, Oldie
If you look at the index methodology you can see that the method is quantitative and based on specific signals from the markets in currencies, commodities and interest rates that are used to determine position size,rebalancing frequency and dealing with futures contracts rolling over. It is a very mechanical process and the index based on it is published daily so the ETF can be tracked against its index. I would imagine a computer system is doing it without human input once initial conditions are set, except to monitor it. The decision to participate or not and how in a market is based on algorithms and the “we” in this case is the algorithmic system as far as I can tell.
@SterlingF: I think the key here is in one of his answers in the first half of the interview: that most of the participants in this market are there to loose money. They are hedging risk not making investments. It makes sense for a wheat farmer to hedge against drought using wheat futures. These ETF’s are just providing the other side of this transaction. It’s sort of like investing in an insurance company. Presumably the active management is using fairly simple diagnostics to figure out what kind of insurance people are buying.
However, this picture doesn’t quite jive with the following two quotes:
“We are probably right 40% of the time, and wrong 60% of the time, but the concept is that the average winning trade pays three units of capital, and the average loss is one unit.” (Isn’t this the opposite of what insurance companies do?)
“About 80% of the liquidity comes from speculators in almost every market. So all we’re doing is participating in a very liquid exposure.” (If the speculators outnumber the hedgers too badly, all you get is risk and no return. Presumably the market is good enough to keep this relatively well balanced.)
@Oldie: I didn’t mean to imply that this was a completely active strategy. It would fall in the “left of centre” on the continuum Dan had mentioned in the comments of his previous post. Passive…but not as passive as a total-market index.
@Andrew: I agree that the decision to participate is based on algorithms and the “we” is the algorithmic system. It’s not a fund manager making the calls but I believe those algorithms would be based on “hunches” to figure out the market trends in the first place.
To me this falls in-line with the Aristocrats methodology. They are both based on rules that they strongly abide to but the rules were decided upon by a group of people and they can probably change at any moment if they really want.
Getting back to what Malkiel said in “A Random Walk…” my underlying scepticism is in the trending analysis algorithm.
@Kiyo: I’m still trying to understand of all this but it makes sense to want to hedge against risk. This is suppose to give you those non-correlated returns people always want right.
This is another example of why this blog is so great. MF’s aren’t really part of the CP strategy IMHO and most people don’t even know what they are….most people (like me) probably don’t want to invest in MF’s anyway BUT it sure pays to be informed.
@SterlingF:
You said:
“They are both based on rules that they strongly abide to but the rules were decided upon by a group of people and they can probably change at any moment if they really want.”
This is the point. While we really don’t know how flexible the manger will be in the future — we are suitably impressed because he’s a professional manager, after all — but the moment he tweaks his algorithm to match the reality of the market he is no different from the nervous investor wondering whether to sell who uses an algorithm to help him confirm that the market is rapidly declining, is overvalued etc etc, a situation that his intuition and experience tell him is true; they are both active managers.
I’m not holding my breath, but my conviction is that they will still try to tweak the algorithms in future.
More to the point, though, I can’t imagine needing to use futures in my portfolio, indexed or otherwise. The moment I start isolating commodities, let alone cotton, West Texas Crude, ethylene, or soybeans as a component in my investments I create such an imbalance destroying the stability an Equity Index/Bond Index mix was supposed to achieve — why on earth would I do that?
I think Dan is teasing us a little, trying to point out that in theory you can have an index for anything, even if it’s of no practical use (at least to CP’s); rather like the philosophy professor pointing out that you can elevate Atheism to be a religion.
SterlingF
I thought, anyone correct me if I am wrong, because of behavioural biases certain indexes do better than others over time – like low beta value and momentum. Even though they should not they do because the biases persist because humans are humans – like the idea that high growth is better – there is a bias in market participants to this which results in the better performance of the lower growth stocks relative to the market. The growth stocks get overpriced consistently so they underperform relative to low growth. Look at the Morningstar value and momentum indexes and methodology and performance over time to see. Dan has written about this.
There is no “objective” index except the economy – so what we are trying to do is capture some aspect of the performance, growth, of businesses in the economy. If the entire market were indexed in the most simple way (ie no one owned anything but the index) then that index performance would in the end reflect the performance of the economy, or at least those aspects of the economy, the businesses, subject to finance within the index.
With respect to the managed futures index, like Kiyo says because the futures are a hedging tool for many if not most the index methodology can have the performance it does consistently.
“Fama and French basically say the commodity futures markets are zero sum game. Not counting transaction costs, the aggregate participant must break even. The aggregate participant is made up of speculators (CTAs,) and hedgers (commodity producers). Many years ago, Prof Robert Merton did a study that found that speculators lost money to hedgers. He even joked speculators are offering a free social service since they pay to take on risk.”
http://investment-fiduciary.com/2012/07/09/managed-commodities-can-counter-volatilitynot/#more-2772
William Bernstein’s thoughts on commodities futures:
http://www.efficientfrontier.com/ef/0adhoc/stuff.htm
A “pitch” to invest in assets that are non-correlated with stocks and bonds is typically based on two premises:
1) The traditional low correlation between stocks and bonds is disappearing and cannot be relied upon.
2) The new asset class (e.g. commodity managed futures) exhibits lower correlation to stocks and bonds and offers higher risk-adjusted returns as part of a diversified portfolio.
The financial crisis of 2008 provides an exacting test of both premises:
1) Index returns from iShares US show that the S&P 500 (IVV) lost 37.00%, while 7-10 year Treasury bonds (IEF) gained 17.97%. In Canada, the TSX Capped Composite (XIC) lost 33.00%, while government bonds (XGB) gained 9.03%. This evidence suggests that low or even negative correlation between stocks and bonds is alive and well.
2) The closest US equivalent I could find to Auspice’s long/short managed futures is the Deutsche Bank Commodity Index Tracking Fund (DBC), which, based on net asset value in the fund’s annual report, lost 30.83% in 2008. This evidence suggests that, in a crisis like 2008, commodity managed futures move in the same direction as most other assets: down.
The financial crisis of 2008 provides a test of non-correlation when it counts the most. Bonds look like better diversifiers than commodity managed futures and at a better price, with a MER of 0.15% for IEF compared to 0.93% for DBC.
Trying to make any conclusions about managed futures based on DBC’s performance is terribly misguided.
Look at how managed futures hedge funds performed in 2008, or how Auspice’s index performed.