A Simple Idea

The annualized return (excluding dividends) of the S&P 500 over the last 30 years is +8%. If by chance one could eliminate all negative months this return would be +26%. Merlin's primary objective is to predict the sign of the monthly performance of the S&P 500. Its predictive accuracy is of the order of 75%. Applied to the last 30 years, this would have led to an annualized return of about 16%, or twice that of the S&P 500.

In addition to its macro signal, Merlin includes a trading overlay ("Batman"). This overlay is only active during periods of market dislocation and produces buy-signals whose expected return is higher than 5%. As the 2020 live results indicate, the contribution of this overlay can be substantial, however, its occurrence is low. For investors seeking a higher frequency signal, I developed "BoB", a daily model to predict the sign of the daily performance of the S&P 500.

Portfolio Exposure

Merlin is a research tool to support market timing decision-making, not a trading system. To keep the message simple, the exposure to the S&P 500 is 0% or 100%. No other assets are traded, the portfolio is never net short and does not use leverage,

A Limited Number of Trades

Merlin was designed for the management of large institutional portfolios. The annual number of transactions is in the order of a few trades per year depending on market volatility.

How to use Merlin?

Merlin can be used for active portfolio management or portfolio hedging. One can use its signals to trade the S&P 500 -using futures or ETFs- or any of the many indices derived from the S&P 500 (Two times leveraged, Top 100, Low volatility, Momentum, Buy-back, Buy-back with low PE...). A comprehensive list of these indices is maintained by S&P Global Ltd. If there is no ETF for the index you wish to use, many trading software allows you to execute proportional trading orders to build up baskets of stocks. Also, this type of popular basket is often available in real-time from banks like Goldman Sachs. If you wish more information about which index to use at any given time, please feel free to send me an email.

Key Differentiating Factors

Merlin is not exclusively based on past price analysis. In addition, Merlin aims to adapt to an ever-changing environment. It is not an indicator but a method of prediction which, at each iteration, builds a new indicator relevant to prevailing market and macro conditions.

I consider the S&P 500 to be almost perfectly efficient. Merlin was built on the assumption that, on a monthly basis, the S&P converges towards an equilibrium price (often) determined (primarily) by economic fundamentals, but the path to convergence remains elusive. A little less so during periods characterized by high volatility (for rates or stocks) or a high volume of equity index options trading.

Is Merlin reliable?

The current version of Merlin went live in September 2018. Since then, its predictive accuracy has been consistent with out-of-sample simulations but there is absolutely no guarantee that it will last. Like all quantitative methods, Merlin suffers from significant limitations (please take few seconds to visit the section "Limitations").