When I tell people that Twitter is one of my best working tools, most of them think I am joking. Most of them do not use Twitter. Years ago, when FB was still a thing for under-65, someone correctly said that on FB you find the friends you have while on Twitter you find the friends you wish for. What is happening (might happen?) to Elon-Twitter is pretty sad but I would say, so far so good…let’s hope 😉

[This is not meant to be one of those stupid disputes like Apple vs Samsung or Oasis vs Blur. I never used Reddit or Discord, for example. I just want to say that Twitter works for me]

This is an example of why I consider Twitter so powerful. @NomadicSamuel is a travel blogger that decided, bless him, to write about ETFs as well. To say the guy works hard is an understatement, in 2022 I think he published 100 posts between interviews and ETF reviews. In just a year, he created a great community around his profile.

A few weeks ago, one of his followers suggested him an ETF-commodity portfolio that, alongside $COM (the ETF I use also in TheItalianLeatherSofa model portfolio), invested in this UBS ETF called $UEQC (the suggested portfolio had a third ETF that I forgot about). I went to check what it was about and…not long before that tweet I was looking for interesting ETFs available for EU investors, how did I miss that? Plus, I have a friend that works in ETF sales at UBS and every time he pops on my LinkedIn I think “who the f&%k buys ETFs from UBS?”…well, most likely, ME!

Some Basics about Commodity ETFs

Commodities, like other financial instruments, have a spot price: how much it costs to buy that asset now (well, technically in 2 days). Commodities, unlike other financial instruments, are..real, physical stuff. Yes, stocks are real as well but to store them you need only an electronic ledger, like a line on an Excel file that says I own X stocks, with your broker/custodian. If I buy oil then I need a tank, ideally fire-proof, to store it. Ok, I do not think I will be ready to replace Matt Levine and his ability to explain complex matters in a simple way any time soon. The point is: storing commodities can be expensive (they might even rot!) and therefore creating a financial asset that replicates the exposure to a commodity spot price is complicated/impossible.

The most ‘famous’ commodity-as-a-financial-asset, gold, represent an exception to the above rule, since you can store it quite easily…even in your mouth (who watched Home Alone these holidays?). Same for all the other precious metals (maybe not the mouth thing).

Commodity ETFs therefore have to replicate indices that represent exchange-traded commodity futures, agreements to buy a commodity at a future date. To provide continuous exposure, futures have a maturity date, index investments have to sell out of their expiring futures and roll into futures with a longer-dated maturity.

Roll Yield

The difference in price between the near-dated future and the far-dated future is called the “roll yield”. It plays a large part in driving the overall performance of a commodity ETF investment alongside the commodity spot performance, especially in longer-then-1-year investment time frames.

Backwardation
Roll yield is positive when the price of the expiring future is higher than the price of the far-dated futures. This can occur when participants prefer to hold a physical commodity rather than a future, perhaps due to short-term supply disruptions. The premium for short-term delivery is often referred to as the
“convenience yield”. In these conditions, the market is “backwardated” and the positive roll yield makes a positive contribution to commodity ETF investment returns.

Contango
Roll yield is negative when the prices of far-dated futures are higher than the price of the expiring future. Markets are said to be in “contango”. The price difference is largely driven by a commodity’s “cost of carry” – i.e. storage, transportation and any other costs associated with physically holding the
commodity. Costs of carry are still there in a backwardated market, but the convenience yield for holding the physical asset outweighs its cost of carry at the time. Negative roll yields are a drag on commodity ETF investment returns.

Since 2001, commodity markets have been in contango 85% of the time and the average length of a period of contango is 132 days. This compares to states of backwardation which are far less persistent at an average of 23 days.

Rolling methodologies
First-generation indices, such as the Bloomberg Commodity Index (BCOM), are a common investment choice in the ETF world. They typically implement a simple “front-month” rolling methodology, meaning they hold a single future until it approaches expiry, then roll into a new future slightly further along the curve, and repeat.

Considering that markets are more frequently in contango, BCOM index performance and those of the ETFs based on it are “punished” by the roll yield. The longer the holding period, the higher this cost compounds on the investor returns. Here is a ‘rough’ visualization: the white line is the Brent ETF while the blue line is the spot price (I hope I took the right data…):

Several index providers offer “second generation commodity index”, products that aim to mitigate the cost of carry; UBS and Bloomberg created the Constant Maturity Commodity Index (CMCI), a mix of futures that reduces the impact of the roll yield by simultaneously investing at multiple points along the curve and rolling a small portion of its investment every day, thus maintaining a constant maturity.

UEQC

The UBS ETF goal is to offer the investor exposure to the carry/roll yield…without building storage facilities. That’s how commodities producers (and traders like Glencore) make some of their profit: they buy spot at X, sell forward at X+Y and store the commodity while they wait for the future contract to expire (obviously the storage cost has to be <Y).

Instead, UEQC takes a short and long position in two indices that contain the same commodities but with different future maturities. It goes “long CMCI and short BCOM“, generating a positive carry on the assumption that rolling CMCI is less expensive than rolling BCOM, while they have the same sensitivity to commodity spot price variations.

[If you are interested in the details, the ETF goes long CMDJCIER, an Index that measures the returns from a diversified basket of commodity futures contracts from sectors such as Energy, Industrial Metals, Agriculture and Livestock. It is designed to provide diversity across contract tenors and as such invests a weighted amount into available CMCI Standard Constant Maturities, ranging from 3 months up to a maximum of 3 years. It rebalances to the underlying weights of the BCOM Index every month. The BCOM Index measures the returns from a diversified basket of commodity futures contracts from sectors such as energy, industrial metals, agriculture and livestock. It is designed to provide exposure to near-dated commodity futures contracts. The purpose of this is to maximize the roll returns.]

Further details on the ETF that are important:

  • it invests in all commodities except precious metals because, as we said, they are very easy to store and therefore their carry premium is very low.
  • it uses a 2.5x leverage effect. This might sound scary for some readers, but given the long/short nature of the investment, is actually necessary to produce meaningful (more than + or – 2.5%/year) returns.
  • it also includes a fixed income return in USD. Since the long and short positions are balanced 1-to-1, the ‘investment’, the funds you put into the ETF, is used as collateral for the trade and that collateral earns interest like a plain vanilla one-week deposit. Now that cash rates are around 4.5%, this return is material (there is no leverage on this, just to be clear).
  • the ETF replication method is swap-based, meaning if UBS-as-the-swap-counterparty (not UBS-as-the-ETF-sponsor) goes bust the investment as well go puff. In other words, this ETF has a relevant counterparty risk, unlike your $VTI or $SPY.

Alternative Risk Premium

In September, I wrote a post about factor diversification. This ETF might be a good candidate as an uncorrelated, to equity beta, carry strategy to put beside fixed income and FX in the same carry sleeve. This strategy also seems to fit the definition of Alternative Risk Premium:

ARP is an investment category consisting of a wide range of rules-based trading strategies targeting returns representing compensation either for bearing risk or behavioral biases among market participants. These systematic strategies span all the major asset classes, trading equity indices, government bonds, currencies, commodities, credit spreads, volatility, and individual stocks. ARP constituents generally share the following three characteristics: (1) clear economic rationale supported by empirical research, (2) persistent risk-adjusted return distinct from that of traditional beta, and (3) liquid (scalable), rules-based and transparent, with a predominantly long-short trading profile.

Does it work?

Here is the historical, also back-tested, performance of the CMCI (white line) and BCOM (blue line) Total Return Indexes:

What is important here is the spread between the two lines, not how every single line performs

And here is the index on which the ETF is based:

It is a great back-test 🙂 note how it becomes more volatile after the ETF went live…coincidence?

Here is the performance of the ETF “live”:

There are some optimistic aspects:

  • the overall, out of sample, performance is positive.
  • drawdowns are pretty shallow
  • gains are concentrated in short periods (not a positive per se but an indication that the strategy is working as intended; if you have a straight, increasing line then there might be some overfitting going on)

Obviously, three years are not enough data to form a definitive opinion but…this is what we have.

Here are some past oil curves, for you to see how the market changed between contango and backwardation in the recent past:

The strategy performs if the ‘far’ part of the curve is less steep than the ‘near’ part when we are in contango and if it is steeper when we are in backwardation.

Why it should be so? As in the FX market, some participants are not primarily driven by profit. For example, they use a particular commodity in their process but their profit margin is driven by other inputs: convenience is more important for them than price. Some participants might have access only to the very near part of the curve because they do not have the ISDA documentation necessary to hedge for longer tenors.

These are just my conjectures. But they are the key aspect to judge if this product can stand on its own legs or if the past performance was pure data mining. The ETF index is rule-based, constructed on fundamental and liquidity data. This is a positive aspect but might also be a weakness because it means those rules can be front-run. Judging if a systematic approach is better than a discretionary one in this case is not easy: a discretionary manager might have a better feeling of temporary anomalies and profit from them (I worked with an Australian manager that does Relative Value trades in the fixed income space and a big part of their process is discretionary).

UEQC and other commodity ETFs

I created a 50/50 portfolio of $COM and $UEQC to check how big is the diversification benefit compared to a 100% allocation to $COM. To set expectations right, $COM had a very strong performance in the last three years (c16%/year): improving from there should be hard.

the white line is our 50/50 portfolio while the orange line is $COM alone

Total Return for our portfolio is lower but we get there with a lot less volatility…which is not surprising if you look at $UEQC equity line. The analysis period is too short to get any concrete conclusion, the real test will happen once commodities enter a bear market (the last 6 months have been just a sideway drift).

A part of me really desires this to work, and it is probably where the danger hides. $UEQC seems like a nice idea but I am not 100% sure the fundamental reason why that “carry” should be there for investors. I hope it is just me being stupid.

What I am reading now:

Follow me on Twitter @nprotasoni