Conviction - Blessing or Curse (or both)?

I found myself pondering why “high-conviction” appears so prominently as a virtue in the marketing material of active managers everywhere.  Why do people place such high value on conviction; why do we abhor the fence-sitters?  “Belief” is such a part of human culture everywhere, as in, “you HAVE to believe in something, right?”

If most active managers proclaim high conviction but more get it wrong than right, clearly some of that conviction must be misplaced.  Conviction is simply a view-amplifier; if you ARE getting it wrong, investors would be better off with less rather than more conviction.  Dogma is dangerous when it comes to investing and these issues are worth investigating.

There has been so much written in the field of behavioural finance that should give us pause for reflection: over-confidence, confirmation bias, herd behaviour, hindsight bias etc.  How do we know that managers are being as clear-headed as they sound; are their assessments of their own skill levels objective?  I prodded and found this bit of research (Plous, 1993), which left me intrigued:

  • Over-confidence is greatest when (forecasting) accuracy is near chance levels
  • Over-confidence diminishes as accuracy increases from 50% to 80%
  • Once accuracy exceeds 80%, people often become under-confident
  • Discrepancies between accuracy and confidence are NOT related to a decision-maker’s intelligence

Curious about how this applies to fund managers, I found a paper by James Montier in 2006, titled “Behaving Badly”.  300 professional money managers were asked to confidentially rate themselves.  An unbiased result would have produced a symmetrical distribution around the “average”; these were the actual results:

  • Below average: 0%
  • Average: 24%
  • Above average: 76%

Clearly some circumspection is required!  However, is it appropriate to castigate active managers en masse?  Attempts were made, most notably by Petajisto and Cremer in 2009, to divide active managers up according to certain traits.  They managed to show that, after fees, “Closet Indexers” underperformed consistently, those that took sizeable “factor bets” (big macro, style, thematic or sector positions) fared the worst, whilst “Concentrated Stock Pickers” evidenced skill before fees, but failed to deliver net positive returns after fees.

The only category of managers to have outperformed consistently after fees where the “Diversified Stock Pickers”.  These are the managers that take large stock positions within sectors relative to the benchmark (active share) but remain reasonably diversified in all other respects.  These results have been challenged subsequently.  An AQR paper found that, when adjusted for benchmark effects, active share had little predictive power.

Equally, it has also been found that holding period is only randomly correlated to performance.  However, in a 2015 draft paper (Patient Capital Outperformance), Cremers and Pareek have evidenced that high active share coupled with long holding periods produced significant outperformance.  Whilst we remain sceptical about single “silver bullet” formulas that predict future performance, the results are intriguing as they resonate powerfully with the attributes we seek in active managers and have applied with material success in our portfolios.  We prefer equity managers that:

  • Take many independent, uncorrelated bets
  • Introduce active risk deliberately – risks created inadvertently create noise
  • Size portfolio positions sensibly
  • Exhibit patience and don’t succumb to short-termism

Concentration IS conceptually the opposite of diversification, but what is definitely NOT true is that they can’t co-exist – in fact they MUST.  The principles are intuitive: concentrate where likely forecasting success is higher and diversify where it is lower.  Skill DOES exist, but the challenge is finding it.  There is never going to be a one dimensional approach that separates the good from the bad, but a pragmatic approach that blends theory with judgement has produced some great results.

[This article first appeared in the June edition of The Trade Press, the publication of the Federation of European Independent Financial Advisers (FEIFA)]