Oct 1st 2017 - by Duncan Smith
Editor's Note: We are thrilled to have Duncan Smith guest blogging for us. Duncan is the longest active member of the investment management industry that we know of with a career spanning well over 50 years. Several of us were mentored by Duncan as young professionals through our association with Russell Investments. Duncan literally wrote the book on investment manager research at George Russell's request in 1996, just before the company sold to Northwestern Mutual Life. There is perhaps no firm out there more influential and pioneering in the manager research space as Russell. They practically invented pension consulting as we know it today. Duncan's wit and cutting insight are perhaps what he is known best for, but I know him best for his generosity and willingness to talk for hours about a topic that is so fundamental to RedQuarry: conducting the best investment research possible over the longest period of time to improve the outcomes for clients.
These rules will help any active manager where success is determined by comparing results relative to those of a benchmark. As relative performance improves, the manager will begin to enjoy more favorable rankings in peer-group comparisons. As will become evident, most of these rules would be difficult if not impossible for the new breed of factor-driven quantitative managers to follow. As the lion hunter wearing running shoes explained to his companions wearing boots, “If a lion charges, I only need to outrun you fellows”.
Your survival kit is a collection of rules – consider them with care and discuss them with colleagues:
Verify that the benchmark proposed by a client or the consultant is a fair representation of the universe of securities from which your portfolios are drawn. It is also important to understand the weighting scheme used to construct the benchmark and how issues are added or dropped.
Identify the “benchmark traps”. These are typically issues in the top 10 or 20 when ranked by their weight in the benchmark. A “We never own…” posture toward one or two of these stocks will guarantee that you will sometimes benefit and sometimes be penalized by their relative performance. The time to make the client aware of this “mismatch” is at the outset of the relationship.
When a manager agrees to be measured relative to a specific index, the short-term comparisons are often dominated by how the manager approaches owning, or not, the 4-5 issues that make up the top 10% of the benchmark’s market value. The manager must realize that it is necessary that they do sufficient research of these companies to arrive at an informed opinion of their relative attractiveness. A neutral view would suggest owning the issue at its benchmark weight; obviously, positive or negative expectations would suggest over- or under-weights. A decision to avoid any exposure to the issues in the top 10% should raise concerns over whether the benchmark is truly representative of the manager’s universe.
Managers of separate accounts need to know whether they have been selected for a specific role in a multi-manager portfolio or represents the only manager for a particular asset class (equity, fixed income, non-US equity or fixed income, etc.). A narrow assignment argues for a “specialized” benchmark (one constructed to reflect a particular investment style or capitalization sector). A concentrated portfolio (~20 positions) is to be preferred for narrow assignments.
Allow winners to run, particularly if the benchmark is capitalization weighted and the issue is in the index. Avoid artificial “position limits” that result in trimming.
Be intolerant of issues that are not behaving as expected. The best place to examine a poor performer is in the locker room not the playing field.
Accept that there are no unique investment approaches. Over a 4-quarter horizon, the probability that your analysis will identify a true opportunity is inversely proportional to the number of earnings estimates posted with the various compilers. As the horizon extends the number of estimates tends to decline, and this can guide in-house analysts where to focus their efforts. As a general rule, active managers will have little variation in the estimates for a given company over the next 1-4 quarters, but very few will invest more than 3%. A conviction high conviction in the 3-4 year estimates might support a 6-7% investment. Over the short run, then, the concentrated portfolio will outperform the broadly diversified competitors and the benchmark.
It is easy to find flaws in these simple rules, and there is no assurance they will somehow replace hard work. A bit of personal history from my Buy Side days might quiet some of the, “But what about” tweets. From 1964 through early 1977, my firm always owned IBM; it was a permanent fixture on the Buy List and was regularly trimmed back when it reached 5% of the portfolio value. Interestingly, for most of that time, IBM represented a little over 5% of the S&P 500. Although we favored the stock, our unwillingness to invest more than 5% resulted in an almost perpetual “short position” in IBM.
RedQuarry's cutting edge research management software is used by leading investment firms around the world to manage their investment process, due diligence files, documents, reporting, and research compliance. Contact us for a demo