We're off to see the wizard, The Wonderful Wizard of Oz.
We hear he is a whiz of a wiz, if ever a wiz there was.
If ever, oh ever, a wiz there was, The Wizard of Oz is one because
Because, because, because, because, because
Because of the wonderful things he does!
“We’re off to see the Wizard” - The Wizard of Oz
Getting to the heart of questions
A debate coach once explained to my class that the winner of an argument is the one who dives deeper into the logic cycles of “why” and “because”. Recalling the lyrics of “We’re off to see the Wizard”, he spoke of having explanations behind each “because” in the song. For each argument you make, you must be prepared for a few cycles of explanation - a few levels of “because”. You should be digging away at the relationships of cause-and-effect until you arrive at a logical premise. Ideally, you should reach a level of argument where the facts are provable by logic or deduction rather than anecdotal evidence or circumstantial inference.
In contrast, investment behavior is almost completely driven by circumstantial inference, trend and pattern recognition, fads, and correlations mistakenly taken as causalities. For example, with a cyclical price graph (where the asset price oscillates over time) an unscrupulous salesman can zoom-in or zoom-out on the time horizon or at different time periods to make contradictory claims about the long term trends and true value of that asset. This data manipulator can demonstrate false trends based on statistics and charts, all without understanding the interplay of true fundamental forces – supply & demand, cost of raw materials, substitutions, negative or positive externalities applied with regulation or subsidies, changing consumer preferences, etcetera – which affect the price of that asset.
There’s plenty of room for biased (or dishonest) investment consultants and managers to misunderstand or misrepresent the reasons for their underperformance. In the previous month’s article, we looked at an investment consultant level – the portfolio level – to consider attribution of individual investments, appropriately weighted to see where underperformance was coming from. Now, let’s go one level deeper and consider the individual investment’s relative performance vs. an index.
Investment level attribution
Continuing our example from the February article, let us say XYZ Large Cap Core Fund is up 15%, while its benchmark, the S&P 500, is up 10%. Why is it outperforming? The reasons are usually interpreted with some very straightforward math. In the February article, we expressed the portfolio’s performance as the weighted average of the underlying individual manager’s product performance. Similarly, we can now consider the individual product’s performance as the weighted average of the underlying sectors and we can further attribute how much relative impact came from weighting of the outperforming and underperforming sectors (i.e. sector allocation) and how much came from picking the right stocks within those sectors (i.e. stock selection).
To explain, Standard & Poors divides stocks into 10 sectors: Consumer Staples, Consumer Discretionary, Healthcare, Materials, Industrials, Utilities, Technology, Financials, Energy, and Telecommunications. Imagine, for our example, the Consumer Discretionary sector stocks are up 20% for the year, handily outperforming the S&P 500 average. Sadly, the XYZ Large Cap Core fund portfolio managers did not predict that the Consumer Discretionary sector stocks were going to do so well, so they actually have less representation (weighting) of that sector in their fund compared to the S&P 500. Specifically, the S&P 500 has 10% of its total weight in Consumer Discretionary stocks while the XYZ fund has only 9% of its total fund in high-flying Consumer Discretionary stocks. In fact, if you look at all the sectors and add them up, the XYZ fund tended to put more weight in the underperforming sectors and less weight on the outperforming sectors, like Consumer Discretionary. Thus, the XYZ Large Cap Core fund had a mildly negative sector allocation effect, which has acted as a drag on their relative performance.
But wait! We said that XYZ was beating its benchmark by a full 5%. How is this possible with a sector allocation drag? The answer is stock selection. XYZ may not have bet on the right weights within each sector, but the stocks that it did pick were far-and-away the best performers within each sector. For example, the best individual stock in the Consumer Discretionary sector happened to be Starbucks; XYZ Large Cap Core fund’s only holding in Consumer Discretionary stock was Starbucks stock. If you add up all the relative winners and losers within their sectors, you see that XYZ picked a lot of the best stocks and avoided the underperformers. Therefore, the XYZ Large Cap Core had a significantly positive stock selection effect, which accounts for the 5% of outperformance.
You can go deeper on this type of analysis. Sectors have sub-classifications – industries – and relative performance can be broken down on this level as well. For example, your fund might have a lot of stocks in the Healthcare sector, but which industry in the Healthcare sector? Biotechnology, Pharmaceuticals, Medical Equipment, or Health Care Services? The results can be sliced and diced at fine levels of detail, but appreciable meaning for investors is usually lost beyond industry level analysis.
The limits of quantitative reasoning
There are generally two reasons used to convince investors to give money to an investment manager: quantitative and qualitative. Quantitative factors, which you can numerically measure, usually boils down to a simple argument: this manager has earned more money than the benchmark in the past and we expect them to continue doing so. Quantitative reasoning looks backward at what has happened, hoping the past will repeat itself.
There are a couple of flaws with relying on quantitative reasoning exclusively. First, there is the caveat found in nearly every disclosure, “past performance does not guarantee future results.” Moreover, if you believe in the cyclicality of investment styles and mean reversion, there is a good argument that today’s relative winners are more likely to be tomorrow’s relative underperformers. In fact, from a strictly quantitative perspective, your best bet might as be a manager with good long term results, but who is deeply underperforming more recently.
Qualitative analysis: the conclusion of investment management attribution
At some level, we have to step away from the exercise of splitting hairs on categories of securities and start considering a different type of question. Why did the portfolio managers pick Starbucks? Why were they underweighting Consumer Discretionary stocks? In short, what was the process for selecting those securities?
Whereas quantitative analysis is backward-looking, qualitative analysis focuses on the immeasurable factors which might provide insights into how a product is positioned for the future. Quantitative analysis generates numbers and metrics – standard deviation, conditional value at risk, alpha, beta, a million ratios and, of course, returns. In contrast, qualitative analysis focuses on words and narrative. Qualitative factors focus on the underlying philosophy of the investment product, the people making the decisions, and the process they use to pick investments. To get an accurate sense of how you expect the product to perform, you will need to embrace both types of data. In our next article, we will dive deeper into qualitative analysis for investment managers.