Friday, 11 November 2011

Are Synthetic ETFs a Time-bomb? A Risk Sharing vs. Risk Transferring Tale

Financial risk management revolves about two options - to share or to transfer risk with/to somebody else. The two approaches are legitimate but they are substantially different in terms of their ripple effects. There are three main types of ripple effects: moral hazard, compounding and systemic. The first – moral hazard – is common to both options, but the second – risk compounding – is specific to the transfer of risk. Two examples are enough to clarify the differences between sharing and transfering.

Imagine a group of friends where one comes up with the following proposition: I just met an actuary who told me that there is high probability that one of us will be dead before we reach 65. So, why don't we join in to begin paying a monthly conttribution to buy a pension to the family of whoever has the misfortune of dying earlier? This form of risk sharing is the basis of any life insurance policy and it makes sense to pay a small amount into a pool of money to get life insurance. Of course, by feeling safer the participants in the insurance pool may become more careless and in the end instead of one death the actuary will come back with an estimate of two of them dying before they are 65. This is the so-called moral hazard of people taking more risks when they feel safer.

Now imagine the same group discussing about buying a couple of stocks in the Dow Jones index and their fear that they may turn out losers. Then one of them comes up with the following suggestion: why don't we pool our money together and buy all the stocks in the index, this way even if a couple of them turns out a loser the loss will be shared by all and we still get a return close to the market return. Actually this form of risk sharing is the basis of index-tracking Exchange Traded Funds (known as ETFs); and it makes sense to pay a small fee to a professional fund manager and incur the costs intrinsic to the small deviation from the index (the so-called trailing cost) to achieve a return close to the index. Of course, the more people adhere to this form of investment the more the index becomes volatile because whenever a constituent stock loses the confidence of investors they will dump not just that stock but the entire basket of stocks in the index. This increased volatility means increased risk and it is also part of the so-called moral hazard.

Moral hazard is inherent to any form of risk sharing but its consequences can be priced in the insurance policy or the fund manager’s fee. However, the alternative based on risk transfer may compound rather than reduce the risk. We may illustrate this process through the following example.

ETFs are one the fastest growing financial products. While there were less than 100 ETF funds a few years ago, now there are more than 4000 ETFs - 1300 created in the last two years – attracting more than $1.6 trillion. Currently the ETF offering covers all asset classes enabling investors to take leveraged or unleveraged long and short positions. Most of the newly established ETFs are synthetic products that replicate the returns of an index without owning the underlying securities using instead swaps, futures and other derivative products.

To understand synthetic ETFs imagine that while discussing the traditional form of risk sharing through ETFs one of the friends comes up with the following synthetic ETF alternative: Instead of incurring trailing error costs, why don't we simply ask my bank to invest the money into an alternative portfolio, including up to 10% in derivatives, and swap its return for a return that matches the exact index return?

However, one of his friends felt skeptical about his proposition and asked: but then we would be transferring the bankruptcy risk in a predefined set of 30 different companies into a single bank and an unknown portfolio which must have a higher probability of going bust? Sorry, I do not believe it is wise to exchange the risk of a trailing cost of less than 1% for a 10% bet in your bank and unpredictable losses in the alternative portfolio. This transfer would simply compound the risk we are already taking on the index.

Guys don't worry! First, we will ask a smart fund manager to run the alternative portfolio. Second, we may hedge the two legs of the swap with a hedge fund capable of unloading the risk to a bunch of investors with deep pockets to support any losses. Third, our bank is too big to fail and would be bailed out by the government in case of trouble. So, trust me, there is no chance that the probability of my bank going bust will be near the 10% implicit in your fear about compounding the risk.

Another skeptical added: but what about the systemic risk? Since most fund managers fail to beat the market why would your bank enter into an agreement which is likely to underperform the market or lose money?

Man, as I told you, on top of a synthetic ETF we may make a lot of money creating and selling a number of layers of asset pools backed by the pool in the alternative portfolio.

Sorry, the skeptical replied: wasn’t that the rationale behind the securitization that led to the subprime crisis and the collapse of Lehman Brothers? I do not buy it, because you will end up using some of those extra profits to bribe regulators and rating agencies into increasing the swap size beyond the 10% limit and by filling the alternative portfolio with lower and lower quality assets. By shuffling around the risks the risk you will not be shared but transferred and compounded. Instead you will build a house of cards doomed to be destroyed in a systemic earthquake.

To sum up, with the synthetic ETFs, there is a real risk of transforming what was an excellent risk sharing device into toxic assets that create a serious threat to the financial system.

1 comment: