Scenario Planning Helps Emphasize All-Weather OutcomesScenarioPlanningHelpsEmphasizeAll-WeatherOutcomes

Have you ever stopped to think about what makes for a good or bad investment strategy process, or indeed whether the answer changes over time?  I am prompted to ask the question firstly by the continuing debate about what the future path of inflation looks like.  There are good arguments, it seems, to be on both sides of the debate: personally, I tend to the more pessimistic camp i.e. inflation will become more sustained for longer, but I can certainly recognize a lot of intelligent commentators taking a more optimistic view as well.  This reminded me of what was probably the last time when there was a sustained change in the inflation environment in the UK -- in the early to mid-90s, which was, of course, in the other direction i.e. inflation and inflation expectations fell rapidly.  At that time, I happened to have some responsibility for Mercer’s investment strategy advice process for clients – of which more shortly - so it provides an interesting window on the role of investment strategy alongside today’s debates.

Before I address the strategy process, I think it’s important to state a couple of things that always matter to remove them from the debate:

1. Timing of changes always matters, but note the emphasis on changes!  If I am moving 10% of my portfolio from Asset A to Asset B, and I choose a time for making the change when the relative price of A to B is 10% higher vs. one that is 10% lower, then I am 1% better off.  That’s a big number.  But the logic only applies to changes i.e. money in motion.  Now, of course, timing is difficult and, inevitably, no one always gets it right, but I would argue that it’s not quite as impossible as popular mythology would have you believe, and that effort put into thinking about timing can have a very high payback.
2. Selection of managers, whether internal or external, to execute the strategy always matters.  “More is better than less” as an old colleague used to say, and the greater the difference between successful and unsuccessful management in the asset class concerned, the more that applies.

Before I come to the strategy process question today, let me expand on my past experience.  In the early 90s at Mercer, we were using a classical ALM (Asset and Liability Modeling) process, but my colleague and I decided that we could improve the process by adding some scenario analysis in alongside, and we created five scenarios, centered on a middle of the road scenario matching our ALM assumptions, but ranging from a couple of different “optimistic” scenarios to what I believe we labelled a “plausible worst case” scenario.  Every quarter we would reset the scenarios based on new current macro and market conditions.

As it happened, the plausible worst case we identified at that time kept being a low inflation one for the simple reason that it was clear that, in that scenario, government bond prices kept rising which meant that some of the key liability metrics kept rising, while other asset prices couldn’t keep up.  Now two quasi-contradictory things kept happening almost every quarter:

1. Our colleagues would decry us for having promoted such a bad scenario to our clients, even as a hypothetical possibility.
2. As inflation expectations started to come down rapidly, in the next iteration of this scenario, we kept having to adjust the fall in bond yields down (and hence rise in liabilities upwards).

Leaving aside what this says about behavioral psychology, I would argue that for clients (and consultants) who accepted the process, the addition of these scenarios enhanced the decision-making process by forcing them to explicitly consider how exposed they were to these adverse conditions and what steps might be prudent to take to protect against those conditions.  Moreover, since overall funding levels in those days looked very rosy, the scenarios made clear that the utility of gains in the adverse scenario was a lot higher than any foregone utility of such defensive actions in more optimistic scenarios i.e. in the conditions where the “insurance” was going to be unnecessary, the client could afford to pay it.

To return to the strategy question itself, I think this experience reinforces my first point that the goal of the strategy process should be to make sure that you end up with an outcome that is palatable in all conditions, and not just in the central scenario.  Therefore, actually the strategy question is not “which inflation scenario is right?” but rather “what do I need to do to protect myself against both a higher and lower inflation scenario?” since I don’t believe anyone can say that either is implausible, let alone impossible.

So, why do I think the aforementioned experience is helpful today?  I think it rests with the power of being forced to consider an explicit scenario.  In particular:

• Being forced to think about what that scenario would mean for each asset class; and separately
• Being forced to consider what the aggregate impact would be at total funding level, and how that triangulates versus more benign scenarios.

Of course, no-one will get these assumptions exactly right.  But I do think that what made that scenario process particularly additive was that it was taking place at a macroeconomic turning point, when it was particularly important to challenge expectations that had become embedded.  None of us can know the future, but I do think the combination of COVID-19 and the massive global policy stimulus response to it means that the probability of us being at a macro turning point is much higher than in steady state, and hence that a scenario process which prevents us falling into a singular view of the future is particularly valuable at this point in time.

My goal in this letter is not to try and provide specific answers, but to encourage you to think a lot harder about what are the right questions at this time, but it’s fair to at least speculate on what might such a process show.  Of course, that will vary from client to client, but my generalization would be:

1. Since so many portfolios are bond and credit heavy, the biggest asset exposure will very likely be to higher inflation coupled with lower growth.
2. Depending on the nature of the liabilities, there may be some relief from corresponding falls in the liabilities.
3. Subject to the answers to the above, considering some sort of explicit allocation to protect against this scenario may well make sense.
4. Since there is already some acknowledgment of the possibility of higher inflation in the market, it’s important to reflect the variability of current prices in both scenario design and any “inflation” allocation itself.

So, what are my overall conclusions?

1. The goal of an investment strategy process, in general, is not to predict the future but rather to make sure that you end up with a strategy that can perform acceptably across the range of different possible plausible outlooks.
2. It is worth thinking about adding scenario analysis into that strategy process at any time, but it can be particularly powerful at a macroeconomic turning point, which there is an arguable case that we face today.
3. The power of the scenario analysis is firstly that it forces you to explicitly address what the implications of each scenario, which helps avoid the behavioral finance pitfalls of failing to acknowledge change.  Secondly, it then calibrates the overall outcomes to help judge where the weaknesses are in different strategic options.
4. On inflation, today’s strategy debate should not be about which view of the future is right, but rather how well prepared are you to cope with both versions.
5. My guess would be that for most clients, if they run through such a process, they will find that their biggest exposure is to a rise in inflation and that this will lead them to consider how serious that exposure is and which mitigating actions will make sense.

So happy scenario planning to you all, and on behalf of everyone at QMA, we wish you a safe and enjoyable summer (or winter to our Southern Hemisphere readers!).