It is that time of year, when strategists everywhere polish their crystal ball in an effort to extract from often meager scraps of evidence a seemingly self-evident truth about things yet to come to life. It has been said that he or she who would foresee what is to come should first reflect on what has been, for everything that happens in the market at any time has a genuine resemblance to what happened in previous times.
The naive approach, of course, would be to simply take the predicted risk of say US Equities, currently at 21%, and profess that by the end of 2019 it will have returns equal to plus or minus 21% from where it is today, or anywhere in between, with a two-thirds probability.
There are (at least) two problems with this over simplistic approach. First, what’s with the Schrodinger forecast!? “As long as 2019 isn’t over, US Equities’ return is both +21% and -21%”!! Second, a 68% probability of anything in between isn’t much of a forecast now is it? Is there a better way?
The Axioma Applied Research team having been set to the task, begun this exercise by defining three macro scenarios driving potential market conditions next year. These Base, Best, and Worst-case scenarios were then modeled as a series of transitive stress tests in the Axioma Risk™ platform to quantify their impacts on markets. The analysis was conducted from a USD perspective and details of each scenario can be found in the paper “Outlook for 2019?” on the Axioma website.
As can be seen from the result summary table above, US equities are expected to rise 15% and 21% in the base and best case respectively, and fall 20% in the worst case. Now it may seem like we only marginally narrowed the range of possibilities from the naïve method, but each of these forecasts is now dependent on a specific scenario with predefined metrics which we can track. The next step is to assign some probability to each one and compute a probability-weighted forecast for each market. Taking our US Equities example, if we assign a 60% probability to our base case and 20% to each of the best- and worst-case scenarios, we get a probability-weighted forecast of +9.3% for next year for that asset sub-class. At the current volatility level of 21%, that translates into a forecasted Sharpe ratio of 0.45.
As new economic data and geopolitical developments are revealed during the coming months, we can adjust the probabilities and recompute our revised forecast for each market and adjust our investments accordingly to ensure that we remain aligned with the highest probability-weighted returns at each rebalancing.
Lastly, we used the MAC Monitor portfolio as of November 9 2018 to replicate the analysis on a sample portfolio. The table below summarises those results. Note that the decomposition below is reported as contributions to the portfolio total and not standalone as in the previous table for markets.
In the base-case the portfolio return would be just shy of 5% next year. In the best-case the portfolio could return over 11%, and in the worst-case, we could lose slightly more than 10% of our assets. If we assign a 60% probability to our base case and 20% each to the best and worst-case, then we get a probability-weighted forecast of 3% return for our portfolio for a Sharpe ratio of 0.32.
We may not be satisfied with a Sharpe ratio of 0.32, but we now have a transparent framework for translating our macro forecasts into expected returns and a way to monitor the change in this forecast based on changing probabilities as new information is digested.
We can also use the output of these stress tests to construct scenario-optimal portfolios to serve as guiding posts in our rebalancing cycles. Should the probability of one scenario now seem all but certain and the ensuing Sharpe ratio reflecting the new probabilities not be sufficient or positive, we can then rebalance the portfolio towards the optimal candidate portfolio to ensure that we remain aligned with our macro views (see chart from the paper below for example).
Disclaimer: As a rule of thumb, the forecasts in this report are subject to a different interpretation of the macro and geopolitical situation around the globe. This is not, I hasten to add, a counsel of despair. Rather, it is a necessary preface to a narrative that has been pieced together from broken shards of evidence. That it is possible to do this has always been, for me, one of the great appeals of this job. I certainly see no reason to apologize for it.