In order to save optimize time (which I do not have), I thought I could browse for new, interesting ETFs available for European investors, meaning UCITS compliant, and then write a post about them. This is a search I do regularly, mostly because my Twitter feed talks only about US-listed ETFs, instruments that cannot be bought by retail investors in Europe.
Whenever I start this process I feel extremely lonely.
The lack of efficient tools to perform the task is astonishing, which makes me conclude I am the only one interested in this type of research? Can it be? I found the screener function of JustETF.com and etfdb.com not really good: with the former you can only deduce if an ETF is UCITS compliant based on its domicile while the latter is basically a tool only for US investors. To add insult to injury, everything that I found has instead a very prominent sustainability/ESG filter; probably this is not the place nor time to go into the rabbit hole of why this twist my guts, Bankeronwheels.com just published a guide on the topic that I strongly suggest you read.
In a dark corner of the Bloomberg ETF search engine I found what I was looking for:
If you are wondering why my $2k/month tool is not my first go-to solution, the reason is two-fold:
- my employer pays for it. I do not want to become 100% reliant on something that I can lose tomorrow if I change job.
- it is extremely user un-friendly (unless you “F1 F1” someone in Malaysia to guide you step by step; somehow my moral compass is still set to ‘solve your non-work shit by yourself’).
How many crypto-related ETFs are listed in Europe?!?
I guess too many (BBG says 123 if you are curious). And they are not even UCITS compliant so…why are they there? Anyway, the main conclusion is that compared to the US, where you can access all sorts of interesting ETFs (Risk Parity, Return Stacking, Trend Following, CTAs….ARKK???), in Europe we get none, zero, zilch.
Shit. I guess this post is done? Maybe I can write about an ETF that I have owned for ages that probably no one is aware of (unless you read this old post). Enter XTXC. This is what I wrote back in 2019:
This ETF tracks the performance of the Markit iTraxx Crossover 5-year Total Return Index. The index measures the return for a credit protection seller holding the most current issue of the iTraxx Crossover credit derivative; the performance of the index is generated by three factors: running yield of the CDS (carry), change in CDS prices and the running yield from the funding component (EONIA rate).
XTXC performs like an insurance company: collects premiums and suffers losses when the insured companies defaults; ETF returns are therefore correlated to the stock market (recessions, defaults and market dips are not fully correlated but rhyme well), but returns are steadier and drawdowns less violent. In the last ten years, the index returned 7% annualized with less than 9% volatility. Future returns will likely be lower since the current negative EONIA represent a drag on the fund performance.
This is the closest type of investment I found to p2p lending, with the advantage of having a liquid market every day and the disadvantage that your counterparty is Deutsche Bank, so default risk is not exactly zero.
Let’s dive a bit more into it.
What is a CDS?
According to Investopedia, a credit default swap (CDS) is a particular type of swap designed to transfer the credit exposure of fixed-income products to another party. To swap the risk of default, the lender (the investor who owns a bond) buys a CDS from another investor who agrees to reimburse them if the borrower defaults. Most CDS contracts are maintained via an ongoing premium payment similar to the regular premiums due on an insurance policy. A lender who is worried about a borrower (the bond issuer) defaulting on a loan often uses a CDS to offset or swap that risk.
Still confused? Let’s say I own a bond issued by the Italian Treasury that pays 3.25%. If I am worried that Italy might default, instead of selling the bond and incurring all related trading costs I can buy a CDS, insurance from another market participant: I will pay the CDS seller 1.31% per year as a premium but if Italy defaults, they will reimburse me 100% the value of the bond I own (assuming I bought it at par).
CDS are typically bought by investors that need protection in the short-term since they are very liquid instruments, more than the bonds themselves. The iTraxx Crossover is a CDS index, think about insurance on a EUR High-Yield bond index. If you are an institutional investor that owns a bond portfolio, liquidating it might take from days to months, depending on the liquidity of each position. If the investor needs protection from further mark-to-market losses tomorrow, let’s say to ‘buy time’ and re-assess its portfolio without a rush, it is way more practical and less expensive to buy the CDS index instead of dumping all the bonds at once.
[The are a lot of technical aspects that make a CDS a good but not perfect instrument. The legal definition of a default is not as clear as you might think and, especially when indexes are involved on the long and hedged side of a trade, the basis risk might get substantial]
Performance
The XTXC ETF makes money from three ‘sources’:
- the running yield of the index (carry): this is the insurance premium the ETF buyer gets when they sell protection to the CDS buyer. In the above example on Italian bonds, it is 1.31%.
- change in CDS prices: The iTraxx Crossover has a 5-year maturity and the index provider issue a ‘new run’ every 6 months (if I remember correctly, I did not check). The ETF mirrors the newest available run, meaning every 6 months the ETF buys back the ‘old’ index and sells the ‘new’ one. This process makes the ETF perpetual, a concept similar to the difference between a plain vanilla bond fund and a target-dated one or what commodity ETFs do. If nothing has changed in the world, the price to buy back the index is lower than six months ago: the risk of default is the same but the insurance covers a shorter period. If a component of the index defaults, or market participants think that the risk of default is now higher (a recession is more likely), the index value is higher and the ETF makes a loss on the rollover. Only a default of a name in the index causes a permanent loss of capital.
- the running yield from the funding component (
EONIA€STR rate). Every swap, CDS included, starts with a mark-to-market = 0. This means that when the ETF enters into a position, i.e. sell the iTraxx, it does not have to provide any capital; the ‘liquidity’ of the ETF (I honestly do not know how to call it) is used as collateral against the swap mark-to-market. This is why it earns a return = EONIA.
When you combine these three elements together, you get a return profile that is similar to a…basket of floating high-yield bonds where also the spread (in this case, the insurance premium) is also floating.
So, a bit like a bond, a bit like a stock; is it better than the two? Is it worse than the two? Does it provide valuable diversification to a 60/40 portfolio?
Portfolio Visualizer
When I first bought it, the graph looked like this:
How many questions do you think I asked myself?
Now that I am older, and wiser, and the ETF returned 0% in the last three years, I decided it is time to make some tests.
Turns out that there is no equivalent to this ETF available to US investors. Not that I particularly care but this means I cannot use either Portfolio Visualizer, either Composer. I found a self-proclaimed Portfolio Visualizer for European investors, Curvo…but it does not work with XTXC either.
Mmmmmmmh. Sounds like it is time to learn how to do it with Bloomberg. Composer really spoiled me, to build a portfolio in BBG is not intuitive at all but in the end I got there. BBG rebalances the portfolio on a daily basis (probably you can set your own timeframe only if you pay $4k a month…) and does not take into account trading costs or slippage, so do not take these results at 100% face value.
First I tried to see how a 50/50 portfolio of XTXC and GLAG (the bond Global Aggregate ETF) would compare against 100% GLAG:
Both ETFs are in EUR, unhedged: that’s why GLAG is almost flat in 2022 instead of being an arrow pointing to the lower right corner of your screen. As you can see from the improved Sharpe ratio, we might be onto something (look at the last two columns, all the rest is irrelevant). Two main considerations:
- if we exclude 2022, the whole back-test period has been characterized by decreasing and low-interest rates. The benefit of XTXC very short duration kicked in only in the second half of this year. The future might look better…or not. We do not have enough data to say.
- The Downside Risk for the XTXC portfolio is better but what happened in the first half of 2020 is reaaaaaly ugly. It invalidates by itself the fact that the portfolio has a better Sharpe, since you cannot lever up that much and then risk being wiped out in 3 months.
Considering that XTXC sometimes acts like a bond, sometimes like a ‘stock’ (when shit hits the fan), I wanted to see the performance of a 50% stock, 30% AGG, 20% XTXC portfolio and compare it with the standard 60/40. Here are the equity lines of the two portfolios:
and here are some stats:
The Sharpe ratio only marginally improves: we lose a meaningful amount of upside to improve on the downside. The possible silver lining is that, again, the back-test covers a period where stocks did extremely well: it should not be surprising that reducing stock weight would reduce the portfolio returns.
I am not sure how to judge the portfolio performance in 2022:
It had the same low as the 60/40, no bueno. But rates in EUR started to increase only in the second half of the year and they are projected to reach 2.8% in 2023; this means a more than 3% boost to returns compared to the back-test period.
Conclusion
Is XTXC worth an allocation? Results are mixed at best.
Given that the ‘insurance premium’ varies, I would think that a quarterly/yearly portfolio rebalancing should help to increase allocation when the premium is high and reduce it when it is low (’cause the ETF price goes the other way).
Looking at correlations, my worry is that XTXC is simply a worse version of an equity index, the same downside with a lower upside:
While the strategy is quite simple, the replication strategy with swaps presents some additional risks, most notably the default risk of the swap counterparty.
What I am reading now:
Follow me on Twitter @nprotasoni