Black swans have a way of making one keenly aware that one’s risk tolerance may not be aligned with one’s investment philosophy. For me this emerged as I watched my portfolios melt away the 2018 gains made during the tax bill euphoria.
I had two main thoughts this week. The first had to do with investment strategy. When I rebalance my portfolio on a monthly basis, am I an investor or a trader? Checking the market daily and making buy/sell decisions clearly makes me a trader. Checking the market every 6-12 months to make adjustments makes me (probably) an investor. But, rebalancing every 33 days makes me both an investor and a trader. Being a trader, I must have an exit strategy resonant with my risk tolerance.
The second thought had to do with risk tolerance. Do I know my risk tolerance and if I am comfortable with my understanding of that factor, are my investment/trading techniques aligned with that risk tolerance? The last week in the market makes these seemingly sophomoric questions come into focus. Let’s start with simply checking the portfolio once a month and using some strategy like Kipling momentum to make buy/sell decisions. Lags for the Kipling momentum model used to provide buy/sell decisions for a set of assets are selected to work best with a monthly update process. Note, however, that the portfolio is open and exposed to growing risk over the 33-calendar period 24/7. This delay is also present in the so-called “protective asset allocation” (PAA) crisis detector or similar crisis detectors that use price rate-of-change (ROC) metrics to measure market swings. By the time the monthly update flags market turbulence a portfolio may be experienced significant losses.
Perhaps one is uncomfortable with that inter-checkpoint risk. One can design and build a computer model that checks each day’s closing and compares this to a stop limit or an alert limit based on the trailing high price. I built such a system and thought this to be a 32-day improvement over the 33-day wait period. This would surely suffice to give me warning and protection against a market downturn. It did not. The sudden one day drop (= flash crash) occurred in real time over one day and by the time my system of alerts began to fire my portfolio based on yesterday’s return, I had already lost more than 2%. Hard lesson learned!
Another solutions are to use diversification/decorrelation of assets, invest in more than 3 uncorrelated assets in a given portfolio, and/or use position sizing to control the portfolio risk. When the swan flies in this mitigates the damage. In a flash crash situation we have time to consider whether to sell or stay the course before our losses become too great. The problem with this method is that in between the infrequent flash crashes one is missing out on return caused by the position-sizing process. Does the missed return over a long period of time exceed the flash crash loss without the position sizing? Backtests show the answer to be, “yes,” but does our time horizon for when we need the funds in the portfolio tolerate the added turbulence of flying without some kind of risk-reduction seat belt?
OK, you say, but you used not only position sizing in the portfolio, you also included stop limits. You were saved during the recent flash crash because the stops fired and you cashed out of volatile assets in time. On this last crash you are looking very smart. But, the back-tests that I and others have run show that over the long haul routinely setting stops for downside protection significantly reduces the return. Whipsaws can be a painful price to pay for stop limits.
What to make of all of this? It’s a reminder that temporal issues and risk are tied together, just as risk and return are joined at the hip. Risk grows with time, and downside protection methods that reduce risk will always reduce the longer-term returns. How much risk tolerance does one have? There is no certain answer to this question, and any method that pretends to protect you from the downside without costing you in the long term should be considered with suspicion.