This weekly commentary is available as a PDF here.
- Risk management is a core focus at Newfound
- “Risk management” most often translates to a conversation about capital preservation – but risk manifests in many ways
- Manager risk is the risk driven by decisions in active management
- We believe that keeping strategies simple helps reduce the potential negative impacts of manager risk
Risk management is core to our DNA at Newfound. It is most visible in the design of our investment strategies where capital preservation is a key goal.
Our external conversations about risk most often revolve around avoiding the market's significant losses. Yet in internal conversations the concept of risk management goes much deeper.
Many of our investors have asked why we keep our strategy designs so simple. For example, our strategies are often highly limited in their investable universe. They also move to short-term U.S. Treasuries, instead of shorting, when our models identify risk.
We believe that success in risk management requires acknowledging all sources of risk. This includes manager risk.
Manager risk is the risk that our active decision making brings to the table. Active decision making is necessary to control market exposure. Yet the wrong decision will introduce error. While perfect accuracy is an unrealistic expectation, infrequent errors are survivable. It is the compounding of errors that can lead to ruin.
We believe that keeping strategies simple helps prevent compounding errors.
Consider the following example of two strategies. Both strategies invest in the S&P 500 and make investment decisions monthly. The first strategy either invests in the market or invests in zero-return cash ("Invested / Flat"). The second strategy invests in the market or short-sells it ("Invested / Short").
A black-box model drives each strategy. The model predicts, with a given accuracy level, whether the market will be up or down in the next month and invests accordingly.
For model accuracy levels between 0-100%, we generate 100,000 12-month samples of market returns. We then turn these market returns into strategy returns and calculate corresponding equity curves.
We see that above a 60% accuracy level, the expected return for the Invested / Short strategy dwarfs the Invested / Flat strategy. This makes sense: with enough accuracy, the Invested / Short strategy can profit in any market environment.
But we want to frame the conversation to be one about risk, not return. So we define "ruin" as anytime an equity curve experiences a drawdown of 20% or more.
We can see that with the more aggressive strategy the probability of ruin sky-rockets as the model accuracy dips below 50%.
To manage risk, we have to assume we'll be wrong from time to time. In fact, we assume our model will go through extended periods of inaccuracy. Taking a more aggressive approach can mean increased return. It also means our incorrect decisions can compound into strategy ruin.
As a firm seeking to manage risk, this is an unacceptable trade-off. So we embrace simplicity.
In Our Models
We rebalanced our Risk Managed Global Sectors portfolio this week.
Energy remains negative and the only sector we are currently entirely out of.
Negative signals in the infrastructure and materials sectors led to a significant reduction in their position sizes as we begin tranching them out of the portfolio. If their signals remain negative, both positions will be fully removed in coming weeks.
With the available capital, we dramatically increased our position in telecom and distributed the remainder evenly among the remaining sectors.
Underlying these allocations are 4 negative signals, 1 neutral signal, and 6 positive signals. With this mix, we expect that a 7% sell-off in global equities would trigger an initial allocation to cash, with a significant allocation being built by a 10-12% sell-off.