Transaction Costs: Keep tabs on price of investing
Commissions are far from the only fees that investors and managers need to take account of when carrying out investment transactions, writes Joseph Mariathasan
At a glance
• There are three potentially significant elements to transaction costs: commission charges, bid/offer spreads and market impact costs.
• Transaction costs do not generally pose a problem to smaller and medium-sized managers with low turnover.
• It is important that managers take into account the availability of liquidity.
Transaction costs are an integral component of any investment and costs for active management are inevitably higher than for passive approaches. Only one strategy can, theoretically at least, incur no transaction costs and that is a market capitalisation-weighted indexation portfolio. But even here, some transaction costs will be incurred through reinvesting dividends and dealing with corporate actions. That is why even actual market cap-weighted index funds can rarely exactly match the index return and most underperform, some significantly so.
Other strategies that are deemed passive, such as smart beta, that do not just use market capitalisation weights can incur considerable trading costs. They are passive only in the sense that no active stock decisions are being made outside of a specific algorithm.
In practice, there are three potentially significant elements to transaction costs that need to be considered: the lowest element is commission charges; of more importance, particularly in illiquid markets are bid/offer spreads; the third and potentially the largest and often least understood component is market-impact costs.
The challenge for owners of capital is to work out how well their fund managers have taken into account these three elements of transaction costs in the development of their investment strategies.
As Nick Samuels, a director in the management research team at London-based consultancy Redington, points out, one cannot just argue that low transaction costs are always a good thing. How much managers will need to trade will depend on the approach being taken. “You have to think of transaction costs from a net point of view. Is this manager giving me something I can’t get from elsewhere and are these transaction costs just a necessity as part of that process?”
Value managers for example, can have long time horizons, epitomised by Warren Buffett, and such managers clearly do not have to think deeply about transaction costs. In contrast, as Samuels says, other styles, particularly anything to do with momentum, by necessity must have higher turnovers.
For many smaller and mid-sized managers with low turnover, such as AXA Investment Managers’ Framlington Equities, transaction costs do not pose a significant problem. It has a turnover averaging about 20-30% a year over a business cycle, as its holding period is three to five years, says Mark Beveridge, global head of Framlington Equities. With about €50bn in assets under management it is a mid-sized firm running global portfolios, so that market impact costs are also minimal. At that level of turnover, what matters is the overall view of the stock over a period of years and not whether it can bought a few basis points cheaper.
Transaction costs clearly become more significant the larger the fund and the higher the turnover. For higher turnover strategies, the extra transaction costs need to be justified by increased alpha generation. That is certainly not always the case. Some higher turnover strategies can generate transaction costs as high as 2-3% a year.
“We have seen managers who have been blissfully unaware of their transaction costs,” says Nathan Gelber, the chief investment officer of Stamford Associates, an independent consultancy. The problem he finds with many managers, is that when asked to estimate transaction costs they invariably only estimate commissions. “They will not have a grasp of the market impact and the bid/offer spreads and they are a multiple of the commissions from a return perspective.” As a result, Stamford, as a firm, is sceptical of managers who claim an ability to take advantage of short-term share price movements as a central plank of investment strategy.
Are larger investment funds able to take advantage of economies of scale? They may be able to negotiate low commissions, but as Gelber points out, large funds have a huge disadvantage as they are likely to create larger market impacts and be subject to wider spreads. He argues that because investment banks have had to withdraw capital from market-making to a large extent, trading today is largely done an agency basis with no one taking on huge positions any longer. Costs have really risen since 2009 as a result of investment banks withdrawing capital from market making. For large institutions, trades are also executed by appointment.
Samuels also agrees that trading costs are higher for larger funds but also sees a more insidious cost. That is the opportunity cost arising from the fact that managers may end up being unwilling to trade at a time when perhaps they should, purely because of the potential market impact of costs associated with trading in large size. For Redington, that is the real concern associated with larger funds and the hurdle of larger associated transaction costs: “We have the view that the bigger the manager, the harder it is for them to outperform,” says Samuels.
The question for Redington regarding transaction costs is whether they are appropriate for the style of the manager. Samuels finds the better systematic managers have well-defined analytics to measure market impact and other associated transaction costs. They are able to capture these effects within their alpha model. The models will not select stocks unless there is value in owning them over and above the associated transaction cost that would be incurred by trading in them.
For example, RAM Active Investments has annual turnover of 300-400%. A fundamental active manager would struggle to keep transactions costs manageable at those levels. RAM claims that its total transaction costs including market impact at day’s close are about 50bp a year for developed markets and about 70bp for emerging markets, says Emmanuel Hauptmann, senior equity fund manager and founding partner. To keep transaction costs low, given such high turnover, does require, as he admits, a huge emphasis on direct and indirect trading costs. Direct trading costs in the form of commissions only amount to 2-3bps in developed markets.
What matters more for the asset manager are the indirect costs of bid/offer spreads and market impact. This requires considerable focus on the actual mechanism of trading. For example, the scaling of positions is done dynamically to try and ensure that the fund does not run into problems arising from illiquidity. What it means is that positions are scaled according to the liquidity available in the marketplace through time. “We don’t wait for weeks or months to figure out that liquidity in a specific name has dried up and we are going incur costs for the fund because liquidity has dried up,” says Hauptmann.
Not surprisingly, Samuels finds fundamental managers are less equipped to model transaction costs. “They usually have internal trading teams that try and do their best to limit market impact and so on,” he says. “But it is not clear that the fund manager who is initiating a trade is necessarily thinking about trading costs to the same degree as a quant manager.” Clearly, though, as Samuels says, that does vary, with smaller-cap managers being very aware of trading costs across markets because they have to be. “I am not sure that large-cap global managers are going to be aware of the market impact of trading in each of the stocks they own,” he says.