In our last post, Show Me the Continuous Improvement, we saw that a growing number of asset allocators are realizing that “behavioral alpha” is the next frontier – and that fund managers who can differentiate themselves on behavioral self-analysis and data-driven feedback on the investment process are more attractive than those who can’t.
But just as there are plenty of portfolio managers out there who haven’t thought about behavioral alpha before, there are plenty of allocators who are behind the curve on how to identify skilled managers. They are still operating on the premise that informational edge is what separates the best managers from the rest – they’re failing to realize the extent to which technology has killed that informational alpha opportunity, and the extent to which it’s creating a new behavioral alpha opportunity.
In the email exchange below*, one forward-looking fund manager (and an Essentia client) challenged this status quo, and we couldn’t help but cheer for him.
* This correspondence originally appeared in Broadside – the quarterly newsletter from Cape Wrath Capital. Some identifying details have been removed.
From: Brian XXXX
Date: 12 October 2017 at 08:38:49 BST
To: Amul Pandya <amul@XXXX.com>
Thank-you for the follow-up call. In response to your feedback request, here is what we took away from the meeting with you and Adam [PM].
The positives first. We think you are doing a lot of things right in terms of fee structure, alignment of interests and portfolio concentration. Fee structures and attitudes towards active management are changing, and XXXX [the fund] is ahead of the curve.
At this stage however, we have chosen not to proceed with any further due diligence; primarily for two reasons.
Firstly, I struggle to see that Adam has any informational edge. We came to the meeting expecting to hear about channel checks, competitor insight, mystery shops, factory gate surveys, expert networks and other sources of proprietary data. Your research, while thorough, didn’t cover anything more than I could have discovered myself given the time and inclination. I was expecting more than this. My view is that you are under-resourced on the analytical side, and so with a bigger fund (and budget) and a team of analysts, perhaps you would be able to correct this.
Secondly, my job is to identify the best stockpickers in each market, but after reviewing your holdings with our equities team, our conclusion is that you have gathered a collection of companies with some or all of the following characteristics: stretched balance sheets, cyclicality, high operational leverage, corporate governance issues, low earnings quality… the list goes on. I’m not sure how best to describe your approach, but it is not stockpicking.
From: Amul Pandya <amul@XXXX.com>
Date: 12 October 2017 at 09:51:02 BST
To: Brian XXXX
Cc: Adam Rackley <adam@XXXX.com
Subject: Re: feedback
We appreciate your feedback. I have copied Adam into this email as I am sure he would like to discuss your comments further.
From: Adam Rackley <adam@XXXX.com>
Date: 15 October 2017 at 17:01:41 BST
To: Brian XXXX
Cc: Amul Pandya <amul@XXXX.com>
Subject: Re: feedback
Thanks for your email. When smart people challenge what we’re doing it’s a great opportunity for us to reflect and regroup. I should start by saying that I largely agree with your observations. But I challenge your conclusions, which reflect conventional thinking.
It’s true that we have no informational edge.
Our data is sourced primarily from public sources, with a particular focus on published financial statements. The depth of our financial statement analysis, and the quality of our valuation framework may give us some degree of analytical edge, but principally we seek to generate returns through exploiting behavioural inefficiencies.
We believe the market is mostly efficient most of the time, and so the pockets of inefficiency that we target are rare. They result from an over-reaction to unexpected events like profit warnings, management change, industry cyclicality, and operational mishaps, which often happen in combination, with a negative compounding effect on share price performance. Opportunities arise where enough of that underperformance is not justified by fundamentals. The best opportunities are at moments of capitulation where existing shareholders ‘give up’ on a company for emotional or technical reasons.
In some cases these inefficiencies are marked by periods of high volatility, where our assessment of approximate value allows us to ‘trade’ in and out of a position. At other times we have longer holding periods that reflect the market’s gradually improving view of the business over time. Sometimes our assessment of fair value is wrong and the share price doesn’t recover. By frequently updating our fair value models, we can review investments with negative-trending fair values, to understand if they are value-traps, or simply recovery plays that we have bought too early.
So, how do we press a behavioural advantage?
Firstly, we widen the timeframe, which helps us to look beyond current market sentiment. Our valuation framework analyses up to ten years of historic data, for the company and peer-group, and we forecast earnings out to a normalised, or mid-cycle point, which is typically beyond the two or three years on which most relative valuations are based. This approach works well with companies facing cyclical or company-specific challenges, but we must be vigilant against building a cyclical investment thesis on a company which is actually in structural decline.
Secondly, we use two forms of feedback loop to identify and adjust for our own behavioural biases. Our ‘qualitative’ feedback loop is based around a debrief that we carry out on every investment, good and bad, to identify any lessons that can be applied more broadly to the process. These lessons are captured in our ‘Checklist of Mistakes’. Our data-driven feedback loop is a system developed by ‘Essentia Analytics’, who use our daily holdings and transaction data to carry out skill-level performance attribution. Here the key risk is in applying the wrong lesson: the next mistake is often made trying to avoid the last one.
Thirdly, we use visualisation and empathy to incorporate conflicting views on a stock. We look at new ideas both from the perspective of existing investors (with the help of our news-flow analysis and annotated share price charts), and also of short-sellers (with the help of our accounting ‘shenanigans’ analysis). Our watch list of stocks, which often face significant challenges, but have not yet become cheap enough for us to invest, is in essence a short-selling list.
I accept that we are under-resourced on the analytical side, but building a team of analysts is not the answer. The traditional analyst/manager structure is inefficient because most of the value in a piece of research is the hard thinking required by the analytical process, not the finished document itself. While some data or fact-oriented aspects of our process are completed by other members of the team, I believe that any judgement-based analysis must be done by the person (or people) making the investment decision.
For these reasons, our resource plan is to be more effective with what we have by leveraging technology. I believe that better decisions can be made with an optimised combination of human and machine-based decision making. For example, upgrading our screening process with machine learning techniques will improve our conversion rate of new ideas into portfolio holdings, while also further developing our behavioural ‘edge’. At some point we might also introduce a second manager (rather than analyst) into the process.
You are right that many of our holdings face significant challenges.
Identifying a good company is relatively easy, while identifying a good investment is difficult. A common mistake is to substitute a harder question with an easier one. People have a natural inclination to look for coherence and causality, and it feels right that a good company makes a good investment. But the returns you make on an asset are a function of the price that you pay. Lower quality companies can make excellent investments if bought at the right price.
Finally, I disagree with your definition of ‘stockpicking’. If stockpicking is buying companies with the goal of maximising investment returns, then we are stockpickers. But if stockpicking is constructing a ‘feel good’ portfolio of the best quality companies, then we are happy to be excluded from the group.
Adam Rackley, CFA