The problem with active funds

Investors should prefer the certainty of index funds which track the index less passive fees, than the hope and disappointment of active funds which, in aggregate, track the index less active fees.

​In this series of articles, I look at some of the key topics explored in my book “How to Invest With Exchange Traded Funds” that also underpin the portfolio design work Elston does for discretionary managers and financial advisers.
 
Clarifying terms
We believe that typically an index fund or ETF can perfectly well replace an active fund for a given asset class exposure.  As with all disruptive technologies, many column inches have been dedicated to the “active vs passive” debate.  However, with poorly defined terms, much of this is off-point.
If active investing is referring to active (we prefer “dynamic”) asset allocation: we fully concur.  There need be no debate on this topic.  Making informed choices on asset allocation – either using a systematic or non-systematic decision-making process – is an essential part of portfolio management.
If, however, active investing refers to fund manager or security selection, this is more contentious, and this should be the primary topic of debate. 
Theoretical context: the Efficient Market Hypothesis
The theoretical context for this active vs passive debates is centred on the notion of market efficiency.  The efficient market hypothesis is the theory that all asset prices reflect all the available past and present information that might impact that price.  This means that the consistent generation of excess returns at a security level is impossible.  Put differently, it implies that securities always trade at their fair value making it impossible to consistently outperform the overall market based on security selection.  This is consistent with the financial theory that asset prices move randomly and thus cannot be predicted .
Putting the theory into practice means that where markets are informationally efficient (for example developed markets like the US and UK equity markets), consistent outperformance is not achievable, and hence a passive investment strategy make sense (buying and holding a portfolio of all the securities in a benchmark for that asset class exposure).  Where markets are informationally inefficient (for example frontier markets such as Bangladesh, Sri Lanka and Vietnam )  there is opportunity for an active investment strategy to outperform a passive investment strategy net of fees.
Our view is that liquid indexable markets are efficient and therefore in most cases it makes sense to access these markets using index-tracking funds and ETFs, in order to obtain the aggregate return for each market, less passive fees.
This is because, owing to the poor arithmetic of active management, the aggregate return for all active managers is the index less active fees.
The poor arithmetic of active management
Bill Sharpe, the Nobel prize winner, and creator of the eponymous Sharpe Ratio, authored a paper “The Arithmetic of Active Management” that is mindblowing in its simplicity, and is well worth a read.
We all know the criticism of passive investing by active managers is that index fund [for a given asset class] delivers the performance of the index less passive fees so is “guaranteed” to underperform.  That’s true, but it misses a major point.
The premise of Sharpe’s paper is that the performance, in aggregate, of all active managers [for a given asset class] is the index less active fees.
Wait.  Read that again.
Yes, that’s right.  The performance of all active managers is, in aggregate (for a given asset class) the index less active fees.  Sounds like a worse deal than an index fund? It’s because it is.  How is this?
Exploring the arithmetic of active
Take the UK equity market as an example.  There are approximately 600 companies in the FTSE All Share Index.
Now imagine there are only two managers of two active UK equity funds, Dr. Star and Dr. Dog.
Dr. Star consistently buys, with perfect foresight, the top 300 performing shares of the FTSE All Share Index each year, year in year out, consistently over time.  This is because he avoids the bottom 300 worst performing shares.  His performance is stellar.
That means there are 300 shares that Dr. Star does not own, or has sold to another investor, namely to Dr. Dog.
Dr. Dog therefore consistently buys, with perfect error, the worst 300 performing shares of the FTSE All Share Index each year, year in year out, consistently over time.  His performance is terrible.
However, in aggregate, the combined performance of Dr. Star and Dr. Dog is the same as the performance of the index of all 600 stocks, less Dr. Star’s justifiable fees, and Dr. Dog’s unjustifiable fees.
The performance of both active managers is, in aggregate, the index less active fees. It’s a zero sum game.
In the real world the challenge of persistency – persistently outperforming the index to be Dr.Star – means that over time it is very hard, in efficient markets to persistently outperform the index.
So investors have a choice.
They can either pay a game of hope and fear, hoping to consistently find Dr. Star as their manager.  Or they can be less exciting, rational investor who focus on asset allocation and implement it using index fund to buy the whole market for a given asset exposure keep fees down.
Given this poor “arithmetic” of active management, why would you ever chose an active fund (in aggregate, the index less active fees) over a passive fund (in aggregate the index less passive fees)? Quite.
Monitoring performance consistency
The inability of non-index active funds to consistently outperform their respective index is evidenced both in efficient market theory, and in practice.
Consistent with the Efficient Market Hypothesis, studies have shown that actively managed funds generally underperform their respective indices over the long-run and one of the main determinant of performance persistency is fund expenses .  Put differently, lower fee funds offer better value for money than higher fee funds for the same given exposure.  This is a key focus area from the UK regulator as outlined in the Asset Management Market Study.
In practice, the majority of GBP-denominated funds available to UK investors have underperformed a related index over longer time horizons.  Whilst the percentage of funds that have beaten an index over any single year may fluctuate from year to year, no active fund category evaluated has a majority of outperforming active funds when measured over a 10-year period.  This tendency is consistent with findings on US and European based funds, based on the regularly published “SPIVA Study”.
The poor value of active managers who “closet index”
“Closet indexing” is a term first formalised by academics Cremers and Petajisto in 2009 .  It refers to funds whose objectives and fees are characteristic of an active fund, but whose holdings and performance is characteristic of a passive fund.  Their study and metrics around “active share” and “closet indexing” caused a stir in the financial pages on both sides of the Atlantic as active fund managers started to watch the persistent rise of ETFs and other index-tracking products.  The issue around closet index funds is not simply about fees.  It’s as much about transparency and customer expectations.
Understanding Active Share
Active Share is a useful indicator developed by Cremers and Petajisto as to what extent an active (non-index) fund is indeed “active”.  This is because whilst standard metrics such as Tracking Error look at the variability of performance difference, active share looks at to what extent the weight of the holdings within a fund are different to the weight of the holdings within the corresponding index.  The higher the Active Share, the more likely the fund is “True Active”.  The lower the Active Share, the more likely the fund is a “Closet Index”.
How can you define “closet indexing”?
There has been some speculation as to what methodology the Financial Conduct Authority (FCA) used to deem funds a “closet index”.  In this respect, the European Securities and Markets Authority (ESMA), the pan-European regulator’s 2016 paper may be informative.  Their study applied a screen to focus on funds with 1) assets under management of over €50m, 2) an inception date prior to January 2005, 3) Fees of 0.65% or more, and 4) were not marketed as index funds.  Having created this screen, ESMA ran three metrics to test for a fund’s proximity to an index: active share, tracking error and R-Squared.  On this basis, a fund with low active share, low tracking error and high R-Squared means it is very similar to index-tracking fund.
Based on ESMA’s criteria, we estimate that between €400bn and €1,200bn of funds available across the EU could be defined as “closet index” funds.  That’s a lot of wasted fees.
Defining “true active”
We believe there is an essential role to play for “true active”.  By this we mean high conviction fund strategies either at an asset allocation level.
True active (asset allocation level): at an asset allocation level, hedge funds which have the ability to invest across assets and have the ability to vary within wide ranges their risk exposure (by going both long and short and/or deploying leverage) would be defined as “true active”.  Target Absolute Return (TAR) funds could also be defined as true active given the nature of their investment process.  Analysing their performance or setting criteria for performance evaluation is outside the scope of this book.  However given the lacklustre performance both of Hedge Funds in aggregate (as represented by the HFRX index) and of Target Absolute Return funds (as represented by the IA sector performance relative to a simple 60/40 investment strategy), emphasises the need for focus on manager selection, performance consistency and value for money. 
True active (fund level): we would define true active fund managers as those which manage long-only investments, either in hard-to-access asset classes or those which manage investments in readily accessible asset classes but in a successfully idiosyncratic way.  It is the last group of “active managers” that face the most scrutiny as their investment opportunity set is identical to that of the index funds that they aim to beat.
True active managers in traditional long-only asset classes must necessarily take an idiosyncratic non-index based approach.  In order to do so, they need to adopt one or more of the following characteristics, in our view:

• Conviction: the ability to show high conviction by allocating to securities with substantial deviation from benchmark weights (high active share)
• Concentration: the ability to create and manage a concentrated portfolio that includes sufficient securities for diversification purposes (for example, a minimum of 30 securities to create the possibility of a normal distribution), but not as many as the index they are trying to beat.
• Cash Limits: the ability and willingness to allocate up to the maximum level of cash allowed to dampen volatility in a risk off environment.

 
Their success, or otherwise, will depend on the quality of their skill and judgement, the quality of their internal research resource, and their ability to absorb and process information to exploit any information inefficiencies in the market.
True active managers who can consistently deliver on objectives after fees will have no difficulty explaining their skill and no difficulty in attracting clients.  By blending an ETF portfolio with a selection of true active funds, investors can reduce fees on standard asset class exposures to free up fee budget for genuinely differentiated managers.
Summary
In conclusion, “active” and “passive” are lazy terms.  There is no such thing as passive. There is static and dynamic asset allocation, there is systematic and non-systematic tactical allocation, there is index-investing and non-index investing, there are traditional index weighting and alternative index weighting schemes.  The use of any or all of these disciplines requires active choices by investors or managers.