Market Tilt or Timing - Is there a difference in forecasting?

At a recent conference, I heard a large money manager say the following, "We do not market time, but we do take market tilts." Unfortunately, no one was able to ask the manager to clarify the difference between tilts and timing. Aren't they both forecasts?

I have used the phrase market tilts, but I am not sure that investors make or understand the distinction between the two concepts, timing and tilts. Although there are subtle differences, clarification of these terms is important. 

Market timing is the adjustment of exposures in a portfolio based on forward-looking expectations. It is a time series forecast. Market tilts will be an adjustment of weightings or exposures based on the characteristics of the asset, strategy, or premia. It can also be a change in weightings based on the market environment regime. Market timing is analogous to trend-following while market tilts are associates with cross-sectional analysis. Both may require some view past behavior repeating itself.

A portfolio of alternative risk premia may start with the assumption of equal volatility or equal risk contribution. There is no view about the risk premia other than they may have different volatilities. A tilt suggests that the weighting of the portfolio may differ from some volatility equalization.

The characteristics of an alternative risk premium may suggest higher returns or risk. These underlying characteristics may suggest a tilt. For example, some risk premia may underperform during different stages of the business cycle. Other risk premia such as FX carry will underperform when all global rates are compressed or there is a significant dislocation in global markets. Still other risk premia may be sensitive to spikes in volatility. Some ARP are more defensive and will do better during economic "bad times". 

These conditional views may be considered forecast but should be grouped differently than a times series performance forecast. Nevertheless, we view that tilts are forecasts and should be give the same level of care and any timing decision.

Premortem for decision-making - Better than waiting for the decision postmortem

All investors and traders want to get better as decision-makers. They are open to learning and improvement, and a natural way to gain this improvement is through reviewing their actions after the fact. The old adage is that we will learn from our mistakes. If you have a thorough review process, you can form an effective feedback loop to ensure future decisions will not be driven by the mistakes of the past.

For a larger money management firm, if something goes wrong, everyone will be called into a room to try and assess what could have been done better. The investment committee will review the decisions to get it right next time through incorporating what was learned into future decisions. Make the trade, do the review, and learn. 

What if we got the ordering of learning wrong? Gary Klein, an influential figure in natural decision-making, suggests an opposite approach. Focus on a premortem. Researchers have found that prospective hindsight is an effective tool for improving the decision process. Image a negative event has occurred and then trace back to the root cause of the event. Before the decision is acted upon, review what could possibly go wrong. Walk through all of the scenarios for failure. The process of finding out what will go wrong will stop bad decisions from happening. There is no need to have a postmortem if the bad decision was never made. Review. Learn. Do the trade.

If a trader wants to puts on a long bond position, the premortem approach would have the trader assume the position went against him and then walk through with his peers all of the reasons for why this strategy went wrong. Did inflation move higher? Did economic growth increase?  Did bond volatility increase? By looking at the reasons for failure, the trader may rethink his actions. 

Klein has also suggested a promortem. Assume that strategy was put into place and everything went well. The decision-maker can walk through the set of events that made it work. This should be a natural part of the decision process for any trader and is a good exercise to provide the narrative for why an investment could go well with some details. A decision walkthrough can reduce the need to review future failure.

Bracketing Fed Funds rates through a known set of models - Getting good probabilistic estimates

An ensemble approach to modeling can be an effective way to get a good idea of the consensus and differences in forecasts on futures moves in the Fed funds. This is an effective alternative to looking at Fed funds futures and options as a market estimate. 

A nice place to show the impact of ensemble modeling is the Cleveland Fed research site which provides quarterly updates of Fed fund forecasts using seven different models. The next update is at the end of this month. The ensemble is developed through using some variations on Taylor Rules that are relatively easy to estimate and have done a good job in the past. 

The approach is simple. The different forecast models are employed with different forecast assumptions. A nice benefit of this approach is that anyone's forecasts can be used with the same set of models as a comparison. It is not the underlying assumptions that are compared but the forecast output through these standardized sets of models. While the Cleveland Fed updates this approach on a quarterly basis, model forecasts against current expectations can be done at anytime. You can call the differences model dispersion or uncertainty. All of these models have rational intuition, so the dispersion across models tells us something about underlying economic uncertainty.

By comparison to the last Cleveland estimates before year-end, we are in a new monetary world. 2018 is ancient history. We are currently between the 25 percentile and minimum forecast. Trading that range is the place of global macro opportunity. Currently, the entire range of forecasts will be lower and likely tighter.  

Non-normal distributions - Assume tail risks are higher than normal measurements

If you assume a normal asset return distribution world and it does not exist, you will be surprised with return performance especially in the tails and unlikely for the better. Of course, when in doubt, the rule of thumb for any sample of return data is to assume normality. Using the central limit theorem is a good starting proposition for any discussion, but it is not where the return discussion should end. The distribution assumptions are a place for danger with decision-making. 

Assume approximately normal, but don't place your faith in normality. Assume stability in the distribution, but don't place your faith in volatility being stable. Most investors know these two assumptions, but still keep the faith with normality. The key question is determining the costs of sticking to the simple approach.

• In a non-normal world, you may be skewed surprised - more of the probability mass either on the right or left-side of the distribution. Negative skew has a performance risk penalty of a strong outlier.
• In a non-normal world, you may be kurtosis surprised - more of the probability mass is in the center but there is a greater chance of extreme tails of the distribution relative to a normal distribution.

The level of surprise is a function of the amount of stretch or shape-bending away from normality. The real world problem comes in the two forms of risk measure: VaR and CVaR. 

VaR will measure the chance of having a market move outside the 95% confidence region. The non-normal distribution with negative skew or kurtosis will increase the chance of having a tail event. The change of a big move is higher than expected. Simulating a VaR through using a normal distribution ensures that you will be at risk of skew and kurtosis. Using a historic VaR will place you at risk that the sample used is not representative of what may happen in the future.

CVaR will measure the average of moves outside the 95% confidence region. This will represent the magnitude of the tail event and whether the size of a big move is larger than expected. A non-normal distribution will have higher CVaR.

The worst case scenario will be a combination of two problems: a greater likelihood of a bad event and the size of those events being larger than expected. Still, a problem is that non-normality is hard to measure. Some non-normality is related to a mixed distribution of different volatilities. These problems increase if there is changing volatility. An increase in volatility that is not measured correctly will increase the chance of VaR and CVaR being too low not from non-normality but from mismeasurement. 

Market risk will fluctuate when there is a greater chance of mismeasurement and non-normality. Look at the current environment. Volatility has fallen significantly since December. The spikes of February and December 2018 are rolling off short-term samples. The big shocks of the Great Financial Crisis are more than ten years old. The assumptions used to measure VaR and CVaR are critical for assessing current and future risk and as we move to the low end of volatility risks of a tail event surprise increase.

Still, there is not agreement on the impact of the problem. Some researchers show that forecasting with the wrong distribution is significant. Some researchers find that skew and kurtosis will have meaningful impact on portfolio allocations and the resulting VaR and CVaR for a portfolio. However, measurement error makes adding complexity an issue. Other state that the impact is a function of the utility used and that only when there is an extreme risk aversion is the tail problem real. 

Our conclusion is that investors should assume that in a low volatility environment the risk of being wrong is higher. Investors face more risk of tail risk mismeasurement. Tail risk is higher when trading nonlinear returning assets, options.  Most importantly, the risks of tail events are not fully calculated with any simple VaR calculations.