Morningstar and others have called for increased transparency around fund trading costs for years. We do so with good reason: our studies and those of numerous academics have shown that fund costs are one of the better predictors of future outperformance available. However, the most comprehensive reported cost figure, a fund’s total expense ratio, does not include a fund’s transaction costs. The predictive power of costs should be dramatically improved if one could account for all of a fund’s operational expenses, including what it costs the fund to buy and sell securities.
To be clear, stated fund returns are accurate—they are given net of costs. The problem is rather in predicting which funds have the best possibility of outperforming their benchmark or peers in the future. If you want to do this well, then knowing that a fund’s all-in costs are not just the 1.50% AMC (annual management charge), not even just the 1.70% TER (total expense ratio), but are potentially double that or more including transaction costs, is helpful. The higher the overall cost hurdle becomes, the harder it is for managers to keep beating a benchmark over the long term.
There has been some recent publicity around this issue, and the IMA sent out a press release on January 27, 2012 stating that transaction costs are already disclosed and suggesting that they are generally not that material nor that important given they are already reflected in fund’s net returns. This was a missed opportunity for the IMA to educate investors on the impact of costs in general and trading costs in particular.
To put some context around the issue, let’s review the components of trading costs. They include:
-- Dealing commissions: the price paid by the fund to the dealer for transacting
-- Stamp Duty: tax paid on trades
-- Bid/Ask spread
-- Market Impact Cost: the cost caused by the impact of a trade on the market for a security, e.g. a large fund may drive down the price of a security it is selling
-- Opportunity Cost: the difference between the price at the time of the decision and the price obtained at execution that is not attributable to Market Impact. If a fund spreads a trade out through time to diminish market impact, opportunity cost can rise if the price moves adversely in the interim.
Of these, only the first two are disclosed in the fund’s documents (further they are only disclosed as pound values rather than as a percent of assets, making it hard for investors and advisers to compare figures across funds). The problem is that these two items can often be the smallest cost of executing a trade. Indeed, for U.K. equities, trading consultancy ITG found that, excluding stamp duty, in Q1 2011 the average commission was just 10.6 basis points, but the average trading costs, still excluding stamp duty but including implementation shortfall, were 54.2 bps (implementation shortfall refers to the difference between the price at decision and the execution price of the trade).
The implementation shortfall, then, was nearly four times the dealing commissions paid. Further, although ITG did not give a breakdown by market cap for the U.K., funds focussing on smaller cap issues generally have larger implementation shortfalls. U.S. equity trading data from ITG, for example, show that Micro-Cap costs have run from 104.6 bps to 545.4 bps from Q3 2008 through Q1 20011, compared to just 23.6 bps to 48.8bps for commissions. For U.S. small caps, IS costs ran from 88.1bps to 137.3 bps in the same period, compared to 11bps to 21.5 bps for commissions.
Given the above, it should be clear that any focus on dealing commissions and stamp duty alone is missing a large portion of transaction costs and that these costs will generally be higher as funds move down the market-cap scale and as their turnover rises. Fund houses know this and certainly the larger ones carefully monitor the efficiency of their traders. We think the industry would do well to educate investors and improve transparency around these costs so all participants can make informed decisions.
