Stock picking in emerging markets: “eagles” or “foxes”?

Eagles hunt over a much wider area than foxes. Several foxes can cover a single eagle’s hunting ground. Foxes know their hunting ground more intimately than eagles do theirs because foxes can see what’s hidden beneath the trees, can smell their prey’s scent and hear even distant prey moving around. An eagle’s eyesight is superb, but it cannot hear as well as a fox and it does not use its nose for hunting.

Using the same analogy, portfolio managers of a broad emerging market equity strategy resemble “eagles”. Portfolio managers of a single emerging market equity strategy resemble “foxes”. The former have a wider “hunting ground” than country specialists. The latter have more information about the particular EM market they manage. Does this help “foxes” to be better “hunters” of inefficiently priced stocks than “eagles”? There are arguments for saying “Yes”.

Intuitive perspective

Which person is better at stock picking in the Chinese equity market?



We bet your intuition pushed you to answer “B”. Your intuitive choice of “B” is based on your natural belief that person in picture “B” is more likely to know the Chinese language and more likely to know the situation in China better than the person in picture “A”. To use our analogy, “A” looks like an “eagle”, “B” – like a “fox”[1].

Knowledge of local language and a particular country’s situation is important for stock picking, according to expert opinions from the largest sovereign fund in the world which has been selecting portfolio managers in various countries for more than 20 years.

“There may be several reasons why our specialist managers in emerging markets have outperformed the other strategies. These are less efficient markets where doing more or better research seems to pay off in terms of better performance, as well as markets where local managers with an understanding of the local mindset and access to company reports and management in the local language are important.” Norges Bank Investment Management from Investing with external managers.

So the intuitive answer is that “foxes” should be better stock pickers than “eagles”.

Intuition helps us make many good decisions every day. However, for complex tasks, it is wise to double-check one’s answer from a rational perspective, as Nobel Prize winner Daniel Kahneman suggests in his book “Thinking, Fast and Slow”. Deciding who is the better stock picker is a complex task.

Rational perspective

We can easily compare the performance of two funds with a similar strategy. A comparison of “eagle” to “eagle” or “fox” to “fox” is straightforward. Things become more difficult when somebody asks you to compare “eagle” to “fox” in the context of a particular equity market. There are two problems:

  • “Eagle” funds do NOt publish their PErformance in a particular country (NOPE problem). Overall performance of a GEM fund is easily available. Performance of the fund sub-portfolios on China, India, Russia or any other country equity market is not published.
  • Limit of Active Weight Size in “foxes” funds (LAWS problem). Suppose a “fox” portfolio manager likes a particular stock on one of China’s equity markets. Let us call it “OW”. The stock has weight of 10% in MSCI China 10/40 index. The “fox” cannot overweight it as there is a 10% limit for a stock weight in the fund[2]. The “fox” cannot therefore earn alpha. Suppose an “eagle” likes “OW” as well. The stock’s weight within the China sub-portfolio of the MSCI EM index is also 10%. The “eagle” can invest 15% of the fund’s China sub-portfolio in this stock. This is a 5% overweight. The share price of “OW” rockets upwards. “Eagle” then appears to be a great stock picker compared to “fox”. However, this is not a fair comparison as “fox” had no opportunity to show his/her skills in this position.

This can work for underweight positions as well. Suppose a “fox” portfolio manager does not like a particular stock on China market – “UW”. The stock has weight of 10% in MSCI China 10/40 index. The “fox” cannot underweight it by more than 10%. Suppose “eagle” does not like this stock either. The stock’s weight within the MSCI EM China sub-portfolio is 15%. The “eagle” can underweight this stock by 15% within the fund’s China sub-portfolio. This is a 5% stronger underweight than is available to “fox”. If the share price of “UW” plummets, “eagle” again appears to be a much better stock picker than “fox”.

This can also work the other way should investment ideas not work. In such situations, “eagles” can suffer a worse impact than “foxes” due to LAWS, which is not fair.

We solved the “NOPE” problem with the help of performance attribution reports by summing a country’s market performance with a manager’s excess return proxy in that country. Let us assume we want to estimate the 2019 performance of a Russia sub-portfolio for a particular “eagle” fund, which has the MSCI EM index as its benchmark. The benchmark for the sub-portfolio is the MSCI Russia index. In 2019, the MSCI Russia index rose by around 50% in USD terms. The performance attribution report for the sub-portfolio gives us an estimate of the gross-of-fees excess return – the sum of the contributions from all the portfolio positions. Let us assume that this excess return was 5%. We add together the performance of MSCI Russia in 2019 (50%) and a proxy of the excess return of the “eagle” fund Russia sub-portfolio (5%) to get 55% – the result for the “eagle” portfolio manager on the Russian market in 2019. We use this slightly complex method as it: 1) solves the NOPE problem more precisely than a more simple method[3]; and 2) provides a good basis for solving the LAWS problem.

We solved the LAWS problem by adjusting the “eagle” fund’s performance attribution results pro rata using the active weight differences available to “eagles” and “foxes”. In the example with the China sub-portfolio discussed above, the “eagle” fund achieved a significant contribution to the excess return from the 5% overweight in the “OW” stock and the 15% underweight in the “UW” stock. Let us assume that contributions to the fund’s excess return from these positions were 150 basis points (bp) each. To make a fair comparison between “eagle” and “fox” results, we need to adjust the contributions from the “eagle’s” positions in “OW” and “UW”. For “OW”, we subtract 150 bp from the overall excess return of the sub-portfolio. A “fox” had no opportunity to get anything from this position at all. For the “UW” position, we divide 150 bp by 15% (overall underweight in the stock) and multiple the result by 10% (the maximum underweight possible for “foxes”). The result is 100 bp. This is the most a “fox” could get from this position. We subtract an extra 50 bp from the overall total contribution for the China sub-portfolio. For this particular example we stop here because the overall adjustment in terms of active weights is zero. We cut an extra 5% overweight and cut an extra 5% underweight.

If the overall adjustment in active weight is positive or negative, we spread this active weight “budget” over the other similar positions and adjust the contributions from them. Let us assume that in the China example we have only one unfair position in an “eagle” fund – the 5% overweight in the “OW” stock. We cut it. If we stop here, we assume that the alternative use of this 5% overweight resulted in contributing zero to the excess return. This is a weak assumption. There were other overweight positions in the sub-portfolio. If the “eagle” portfolio manager had the choice of somehow redistributing the extra 5% overweight, he/she would increase the overweight in some other stocks that looked attractive. Hence, we redistributed this 5% across all the other overweight positions in line with the initial size of their active weights. Then we adjusted the contributions to the excess return from these positions proportionately. We use the same method in cases where the overall active weight reduced “budget” is negative. The only difference is that we do it with the underweight positions.

The result of our two months’ work on this is shown in Graph 1 on the next page.

In 6 out of 7 large EM countries, “foxes” alpha was higher than that of the “eagles”.

Graph 1. Alpha vs. relevant index for “eagles” and “foxes” who manage funds with 4-5 stars from Morningstar   (in USD terms, annual average, gross of fees)

Note: We calculated Jensen Alpha. For country focused PMs, we used annual returns for 4-5 stars share classes of the funds from a particular category and returns of the relevant MSCI 10/40 country index adjusted for the average fees for these funds’ shares in the category. For GEM PMs, we took a portfolio of eight equally weighted GEM funds from the 4-5 stars group. This portfolio has results very close to the dynamics of the average for 4-5 stars GEM funds group – tracking error vs. the sample average of 119bp, average annual difference of close to zero and similar average ongoing charges during 2010-2020. We used Bloomberg’s performance attribution module of the PORT tool for each year for each country and for each fund to calculate the total effect vs. iShares MSCI EM UCITS ETF for a particular country sub-portfolio. We took the average total effect for the eight funds to get the final estimate of GEM PMs’ gross of fees excess return. We added to the excess return performance of the relevant MSCI country index in order to get an input for alpha estimations. To make the LAWS problem adjustment, we took the performance attribution for the average portfolio for the eight GEM funds each year. We calculated the maximum possible active position on the basis of 10% maximum limit of a stock position in the portfolio and the weight of a stock in a particular MSCI country index (which we got from the iShares MSCI EM portfolio). For Russia, the same procedure was based on MSCI Russia 10/40 index data. For 2020, we took the period January-October and assumed this to be an annual figure. We took the 10 countries with the largest average weights in the MSCI EM index over the last 10 years and selected those for which we had data about 4-5 star funds from Morningstar’s database. We took the period 2012-2020 given that for many “eagle” funds there is no relevant performance attribution data in Bloomberg before 2012. Sources: Morningstar, BNPP AM, Bloomberg, TKB Investment Partners, November 2020

In 6 out of 7 large EM countries “foxes” alpha was more stable than that of the “eagles”.

Graph 2. Volatility of alpha for “eagles” and “foxes” who manage funds with 4-5 stars from Morningstar  (in USD terms, based on annual data, gross of fees)

Note: We used the same data as for Graph 1

Sources: Morningstar, BNPP AM, Bloomberg, TKB Investment Partners, November 2020

Brazil was the only market where “foxes” and “eagles” had the same alpha and where it was less stable for “foxes”. Brazil stands out mainly due to large cash positions in some of the funds in 2016 and 2020 due to large flows. Specifically, for one of the funds there was 50% rise in assets during February-April of 2016 due to inflows. Its stock price rose by almost 40% in USD terms over that period vs. close to 55% for the MSCI Brazil 10/40. The same fund stock price dropped by 9% during one of the weeks of greatest market panic in 2020 (week ending 20 March) vs. a more than 20% drop for the fund benchmark. During the four preceding weeks, the fund lost around 7% in assets due to outflows and was likely preparing for more. “Foxes” got 3.6% alpha vs. 1.8% for “eagles” while the alpha volatilities were 2.3% and 4.6% respectively when we exclude 2016 and 2020 from the analysis.


The most skilled “foxes” delivered larger and more stable alpha than the most skilled “eagles” in all seven of the largest emerging equity markets in our analysis. Country focus and knowledge paid off. This forms a good background for a fund of funds approach to managing portfolios in the emerging markets.


Authors: Egor Kiselev PhD, Head of International Business & Investment Marketing; Gennady Sukhanov, Deputy Head of Equities & Head of Research

[1] The rational answer is – there is not enough information to make a decision. Person on picture A can have 20 years experience analyzing Chinese stocks, while person of picture B can be a CEO of a large solar panel producing company during the same 20 years.

[2] Country specific funds in our sample have 10/40 diversification rule – 10% limit for securities of one issuer and sum of weight of securities with more than 5% weight should not be more than 40%

[3] It is possible to take total return figures from the performance attribution report instead of the total contribution to excess return. This is not very accurate given that Bloomberg’s performance attribution reports are constructed on the basis of fund portfolios which are not available every day

Stock picking in EM eagles or foxes January 2021
Previous Story

Russian Equities Weekly 4-8 January, 2021: Off to a good start

Next Story

Russian Equities Weekly 11-15 January, 2021: The Ministry of Finance restarts FX purchases