In the previous two articles I described how to use pair trading – in our case double short – to benefit from inefficient ETF. We said that when selecting suitable candidates, it is necessary to take into account the borrowing fee (in % per year) for short, the volatility of these markets and, last but not least, the liquidity – the number of shares available for short. While keeping these three principles, some risks need to be considered. I would like to devote them this article.
So, what are the biggest risks?
- Losses are hypothetically unlimited – although we are not shorting single market but usually we short two markets going inversely against each other (expressed in % on a daily basis), the risk is hypothetically unlimited. As described in the previous article, zero volatility is the biggest “killer” of this strategy. If the market goes in one direction, with minimal corrections, the shorting of the second ETF in the pair is not strong enough to balance the losses. To some extent, of course, rebalancing can help – the adjustment of the nominal values on two markets to the same value, but the long-term shortening of the growing market with zero-fluctuations is still a loss-making issue. This situation is shown, for example, on the ETF pair: TNA and TZA, see the following chart. While the TNA (blue curve) has increased by about 137% over the last 2 years, the TZA (red curve) weakened “only” by 76%, so it would have provided insufficient coverage against short-term TNA losses.
- This is not a market-neutral strategy – even though we are rebalancing the position (with a firmly defined “deviation of the nominal” – eg by 10, 20%), it is clear that in the “interval” between the rebalancing, we have a position either slightly long or slightly short. This is not a market-neutral strategy, and we need to be aware of it.
- Transaction costs – on one hand there is the cost of short positions, but it is also necessary to count the costs (commissions) for the opening of the trade during regular rebalancing. In assessing the strategy’s performance these costs must also be taken into account and whole thing must be ideally traded with a broker who has the cost (commission) as low as possible.
- Gain vs. risk – these strategies are usually characterized by a high rate of success, but with a low profit on the other hand, and there is a (low) risk of big loss. Even though they are not option strategies, they are often similar to the profile of the so-called market-neutral option strategy (such as Iron Condor with a very wide profit margin or OTM vertical spread). Also, there exist a high probability of success with a relatively low yield, on the other hand there is small probability of a relatively high loss. Those of you who are engaged in option trading of such strategies will find some parallels with the ETF pairs.
- This is not an opportunity for arbitrage- as it has been said above, losses can simply be very high. Even though our goal is to short ETF with significant inefficiencies (which are pushing the ETFs towards zero in the long run), nobody can guarantee that an ETF will not be growing without a volatility causing us some troubles. In that case, we will not be saved from losses even when we perform a short of the inverse market in the pair. By this we can mitigate but not eliminate our losses.
Despite of these risks, I consider the short of the ETF pairs an interesting trading approach that has a place in the trading portfolio of a trader who is benefiting from inefficient ETFs – for example, because ETFs have no options available, etc. Increased attention should be paid to the principles mentioned above, especially it is important not to see this strategy as an opportunity for arbitrage.