Is high turnover inconsistent with being a long-term investor?
High portfolio turnover is not necessarily a bad thing; it’s important to take a pragmatic approach.
Unstructured Learning Time
It is a truth universally acknowledged that high turnover in an actively managed portfolio is a bad thing.
The inclusion of “actively” in this statement is deliberate, as frequently rather high turnover in a number of passive/index/smart beta products (and the increasingly opaque trading strategies behind many ETFs) appears to get a free pass. Not to mention the frenetic turnover in so-called hedge funds.
Tall tales of turnover
The mainstream long-only fund management community appears to be under the greatest scrutiny by regulators, the commentariat and, by extension, the man in the street. At its most distorted, the impression is given that the average fund manager is as reckless of budgetary considerations in their portfolios as they are in ordering their Notting Hill kitchens.
In itself, it seems a grotesque exaggeration. It also makes little sense. What credible reason is there for a fund manager, judged by results in a very transparent market, willingly to harm his or her performance by unnecessary trading?
This is particularly so given that the scope for funding research through commissions is increasingly constrained and has become a very small proportion of the total cost of trading (spreads, frictional costs, stamp duty as well as conventional brokerage).
Most fund managers are acutely aware that unnecessary turnover impoverishes their clients to the enrichment of many an investment banker and trader.
Turnover and stewardship
There is, however, an important extension to the turnover debate, and that is where it links in with stewardship, and the virtues of long-term investment. To quote (rather selectively) from the Kay review on UK stockmarkets: “ ‘Trading’ is about allocating portions of the cake, while ‘investing’ is about growing the size of the cake overall”.
There is something distinctly pejorative in the implication that high levels of trading, or even trading full stop, are inconsistent with a long-term view.
Intertwined (arguably the inevitable third pillar of this consensus) with this vision of low-turnover investors as careful and virtuous stewards of capital must be the belief that equity valuations rarely deviate materially from some assessment of intrinsic worth.
QED, low turnover can only be consistent with the fiduciary duties of fund managers if deviations from intrinsic worth are conveniently rare in incidence and in extent. This is a position conveniently aligned with the regulatory preference for passive over active, and low cost of ownership over high, even to the lengths of regulators seriously considering an arbitrary cap on fund expense ratios including transaction costs.
It will be unsurprising to hear that, as an active investor, I do believe that share prices frequently and substantially deviate from long-term intrinsic worth. Furthermore, I believe these divergences are exploitable to add value through active management, as long as the appraisal of worth is based on a disciplined assessment of the long-term potential of investee companies.
However, despite this belief in the necessity of longer-term assessment of worth, I also contend there need be no causal connection between having a long-term investment style and low turnover. In other words, low turnover may be a by-product of a long-term investment horizon, but it is neither a sufficient nor even a necessary condition of it being embedded in an investment process.
Churn and burn?
My colleague, Duncan Lamont, has recently done a study (“Why Churn is Not Necessarily Burn”) which shows that there is no clear correlation between low turnover and the likelihood of added value among a universe of US equity funds. Essentially, high turnover managers in the US do as well as low turnover.
Doing a similar exercise across a representative portfolio for 68 different strategies at Schroders over a ten-year period yields the same result. If anything, there is a slight tilt in favour of higher turnover funds, although the latter should be set in the context that a “high turnover” fund in a Schroders context has a holding period of 1 to 1.5 years.
How this all relates to my investment process
A number of portfolios I run, one of which for more than 20 years, fall into this high alpha/”high turnover” bracket.
Etched like Brighton Rock throughout our Asian equity process is a commitment to identifying good companies based on long-term insights. Our analysts are asked to focus closely on sustainability, corporate governance, strategic direction, and the durable moats of an investee company’s business.
They are asked to formulate their fair values based on detailed modelling as far out as they can; this obviously varies by company, but is typically around three years out. On this they postulate a reasonably attainable target price reachable on a 12-18 month basis (i.e. when the broader market may be recognising the potential we have already identified because we look out further than the market).
We constantly review, and if appropriate amend, that fair value, and if such amendments mean there remains sufficiently attractive upside we are happy to continue to hold. As a result there are several stocks I have owned for at least the last decade.
Markets change; our philosophy does not
This has been a tried and tested philosophy over many decades, and certainly throughout the 20 year-plus life of the fund I have run the longest. And yet, an outside observer might be shocked by the extent to which the levels of turnover has varied over the years, as illustrated below.
Variable turnover over long term
The key is that our investment philosophy has not changed, but the nature of the markets has varied through time. Inconvenient as it may be for the proponents of “perfect markets”, in reality equities go through periods of high dislocation and stock price dispersion. What is interesting is how that interacts with a disciplined long-term investment process.
As shorthand for this scope for great variations in valuations within markets, we would look at cross-sectional volatility (i.e. the dispersion of return of all the stocks in the MSCI Asia ex Japan index) of one-year returns.
Source: Schroders, July 2017
Apparent to the naked eye is an immediate coincidence of high turnover in the period from early 2008 through to mid 2011, and a sharp rise in cross-sectional volatility. This was not dictated by some kind of short-term trading frenzy, but by the fact that relative valuations between stocks we held and other stocks not held changed to such a degree relative to long-term intrinsic worth (our analysts’ fair values) that a high level of portfolio activity was necessary. Indeed, it would have been a dereliction of our fiduciary duty to have sat on our hands.
The GFC and beyond
For example, through the Global Financial Crisis in 2008, a very narrow group of very defensive stocks held up while everything else sold off in the general panic. Perhaps, in retrospect, we were not sufficiently defensive going into the crisis, and late 2008 was a poor period for our relative performance. But we stuck to our guns, selling what defensives we had to buy more of what we felt to be the long-term winners thrown out with the bath water.
Dispersion remained somewhat elevated through to the end of 2011, and so did our level of turnover. The three years from end 2008 also coincided with one of the strongest periods of outperformance our history.
However, equally of interest are the periods of elevated dispersion that did not elicit any response in our trading. Most recent was the spike in dispersion in late 2014/early 2015. This was largely a function of the extraordinary Shanghai bull market fuelled by permissive credit, high savings, grandstanding from the leadership, and unprecedented levels of margin finance. We struggled to find attractive stocks in that market, and we sat on our hands as the bubble expanded.
The impact on our relative performance is all too apparent, but when sanity re-asserted itself (amid widespread listing suspensions and the collapse of many ramped stocks including one solar company which reached a peak valuation of over $50bn), our relative performance recovered dramatically.
Dispersion (and turnover) on the up
And what of the last six months or so? Well, dispersion has picked up, and so has our turnover. Many of our holdings, particularly in the information technology sector, have done very well, and so has our relative performance. We are beginning to see more compelling long-term value in other areas, and so we trade…
It seems to me the message is clear. In itself, turnover is not a bad thing, and neither is it necessarily indicative of investment time horizon. However, an arbitrary limit on transaction costs and, by inference, activity, is a dangerously one-dimensional approach, and gives rise to fundamental conflicts of interest for fund managers doing their fiduciary duty.
 With apologies to Jane Austen.