To Top
Close

Manager Selection

Context

Whether private investor or institutional, the question of manager selection is critical. These can be ETFs, active managers, hedge funds, etc. The question is multi-folds:

  • Portfolio construction, management of exposures
  • Selection of the best managers to achieve a desired exposure
  • Risks and diversification
  • Identification of manager risk, frauds, etc.

Both quantitative and qualitative scrutiny are necessary to achieve these goals. Using simplistic quantitative methods have proven not only inefficient but dangerous. Only relying on qualitative analysis to complement simplistic quantitative methods is not enough.

Proposition

Advanced risk and return analysis using Polymodel theory to answer the following questions:

For each single potential manager:

  • Identification of risk sources (hundreds of potential risk factors)
    • Equities
    • Fixed Income: yield curve, relative value
    • Credit, securitized products
    • Commodities
    • International, FX
    • Emerging markets
    • For all asset classes: indices, asset specific, long/short, event driven, options
  • Scenarios that can be damageable for the fund
  • Extreme risk, correlation breaks with markets, hidden risks
  • Fraud warning, suspicious return series
  • “True” alpha, stripped out from extreme risks
  • Skill vs. luck, identification of return sources

For the global portfolio of managers

  • Identification of risk sources (hundreds of potential risk factors)
  • Scenarios that can be damageable for the portfolio
  • Extreme risk, correlation breaks between constituents of the portfolio, hidden risks
  • Diversification assessment, allocation optimization accounting for extreme risks
  • Identification of risk concentrations and of duplicates of risk profiles

Strategy Simulation Assessment

Some funds, whether experienced or emerging managers, propose new investment vehicles based on quantitative simulations. Identifying whether these simulations are likely to maintain their risk/return profile in real life is the investor’s question.

Polymodel analysis is a powerful tool to identify the sources of risks and returns and evaluate the potential of a strategy from its simulations. It allows to concatenate analyses on simulations and on actual trading periods and extrapolate, from the purely simulated portion, the risks of a actual trading portion, even if it is rather short.

Contrary to common practice of discarding simulations non supported by actual trading, polymodel analysis leverages on the full information provided to extract all the information necessary to measure the expected returns of a strategy and how it can fit in an existing portfolio.

Nonlinear Factor Analysis

A polymodel is comprised of a collection of single factor analyses, each of them, individually, being a partial view of the manager risks, but collectively representing a very complete view of the global risk profile. Each single factor analysis provides the – possibly nonlinear – response of the manager to a move of the factor, thus capturing correlation breaks that occur under unusually large factor shifts. Over the long run, the convex (i.e. antifragile) or concave (i.e. fragile) response comes with a cost, or a profit, which should be accounted for in the evaluation of the fund alpha with respect to the risk factor.

The Beta  of this Long-Short Equity Fund changes in a systematic manner. The traditional linear Beta of the fund is 0.6. But by carefully examining these returns, one finds that the “upside Beta” is 0, while the “downside Beta” reaches 3. Had the S&P fallen by 15% instead of 5%, the fund would have lost an extra 30%, reaching almost 40%, sufficient to put it out of business.

Performance of a Portfolio of Hedge Funds Selected by Polymodel

We present here the performance of a monthly rebalanced portfolio of hedge funds that are in the top 10%, both by Long-Term Risk-Adjusted Expected Return (LT-RAER) and by Sharpe Ratio, with Sharpe > 1. In market rally, the portfolio fully captures the upside, but LT-RAER protects the downside in the downturn.

The HFR database of monthly hedge fund returns was used for this simulation. To account for redemption delays, portfolio actual rebalancing is effectuated 3 months after the funds are ranked and selected.