A new study by Lumint Corporation and New Change FX (NCFX) argues asset managers are “turning a blind eye” to FX costs and passing these costs directly on to end investors by means of a lower net asset value. The study finds the average transaction cost of an overlay FX hedge is $267 per million, something that results in a “substantial underperformance across the board” by currency hedged share classes.
“These are soft dollar arrangements in everything but name,” NCFX argues. “Hedged share class investors are subsidising the unhedged master class investors of these funds. The only way to control this is to conduct proper, independent, transaction cost analysis.”
More than a decade ago a number of US states sued State Street and Bank of New York Mellon over the standing instruction cases, gaining substantial settlements – often the custodian bank executes what are largely a large number of small administrative trades through a back office function. The paper by Lumint and NCFX argues this means that share class hedges, which are similar trades but not the same, “are not necessarily traded with the same diligence that front office deals might receive”.
The study selected 30 of the largest asset managers selling funds in Europe with a USD reference or master share class and focussed on share-classes hedged into euro, Sterling and Swiss francs. Returns were measured from January 1st, 2018 to September 1st 2020, and any hedged class with fewer than 12 months of track record was excluded.
“In our experience there has been a tendency amongst investors to dismiss FX costs as being insignificant, or worse, ‘coming out in the wash’.”
Then, using manager submitted monthly return data from Eikon, the study identified the non-base currency sub-funds and mapped each sub-fund to its USD master share class, ensuring that distributing share classes, institutional share classes and so on were correctly mapped. This enabled them to identify the monthly gross performance difference between the hedged and the unhedged share classes of fund. The underlying investment strategy driving returns is identical in the two mapped funds, but the hedged share class also reflects the return of the hedge. Notionally these two-return series should closely match, the firms say.
Having established the gross performance difference between the funds, Limint and NCFX then considered the cost of the unavoidable aspects of hedging a portfolio and removed the performance of the interest rate differential using the regulated NCFX interest rate differential benchmark. It then accountted for the effect of fee differentials between hedged and unhedged classes and the effect of future value drift. Future value drift reflects the effect of hedging forward based on a present value, so the future value is unknown when the hedge is placed. This then leaves an element that is attributable to FX transaction costs.
Finally, and only where the manager has submitted monthly AUM data, the study created an estimate of trading volume in order to relate the transaction costs identified through the performance metrics to a dollar cost for hedging the portfolio. This captures the movement of cashflows in and out of the sub-fund.
A volume element related to a re-hedging model based on limits of 101% and 98% was then added, should the currency move outside of these boundaries re-hedging occurs and the trading volume increases. “This is a very tight re-hedging parameter, so we would expect to be over-estimating volume and therefore possibly underestimating costs,” the paper states. “This is therefore a conservative estimate of transaction costs.”
In a stark performance difference, the study finds that the hedged portfolio cumulative return for the period of January 1 2018 to August 31 2020 is 3.16% , whereas the return for the unhedged USD reference classes is 11.84%. “The performance pattern is clearly offset by a cumulative drag over time on the Hedged Share Classes,” the paper observes. “This indicates that the costs incurred are structural and significant over time.”
Recognising that any currency hedging involves a degree of cost, the paper seeks to ascertain whether that cost is in line with reasonable expectations. It removes unavoidable costs such as the interest rate differential, fees and “a component” for future value drift, to leave what it argues might be direct FX execution costs.
The paper was able to identify AUM for 19 of the 30 managers resulting in total trading volume of $1.6 trillion over the almost three-year period and a cost of $432 million – or $267 per million. The majority of the transactions were monthly rolls. Against this, NCFX says the average cost reported by its TCA analysis for swap transactions is $35 per million, with the best traders expecting a cost in the region of $4 per million.
In the case that these transactions were conducted at the NCFX average price, then the total cost would have been $56 million; a saving of over $370 million, or an average annual improvement to return of 29bps, according to the paper, which adds that reaching the best price of $4 per million traded would see a total trading cost of $6.4 million over the 32 months surveyed. “The trading costs identified here are on average more than 10 times higher than they should be.,” NCFX says, adding, “If we take the UK asset management industry’s assets of $10 trillion and assume that 24% of them are invested overseas, as per the ratio of domestic to hedged foreign investment identified here, then the annual cost from trading is in excess of $5 billion per annum. Given the exceedingly low returns available, mismanagement of FX hedging is costing investors a large portion of their returns.”
The paper argues that investors need to assess the control process for managing FX costs within the business responsible for share class hedging, specifically the incentives driving the actions of those responsible for running the currency hedging programme. “For instance, when this is the asset manager themselves, it is clear they should be incentivised to minimise FX costs for their funds,” the paper states. “But what if the outsource provider is the custodian bank for the fund, acting as principal and executing the deals with their own FX trading desk?
“Here the incentives are clearly to charge higher execution costs where possible,” it continues. “To address these inherent conflicts of interest, investors and asset managers should always have transparent and specific pricing arrangements in place as well as oversight procedures to review and enforce those arrangements.”
The paper further argues that asset consultants must begin to fully consider FX costs in their investment recommendations, the firms observe, “In our experience there has been a tendency amongst investors to dismiss FX costs as being insignificant, or worse, ‘coming out in the wash’.
“Asset consultants have tended to focus on the Total Expense Ratio at the expense of the implicit costs paid through trading,” it continues. “This combined with a casual approach to conflicts of interest inherent in selecting conflicted Transaction Cost Analysis and the agency versus principal relationships within many dealing arrangements means that investors are overpaying for FX services.”
The paper concludes by stressing the need for asset consultants to help investors manage their FX costs lower. “Above all, investors should ensure they avoid conflicts of interest where possible and closely manage them where they are impossible to avoid. To do so, investors should utilise independent, regulated TCA benchmarks to oversee all outsourced activity. Using effective TCA for oversight, regardless of whether they use an independent agent or their custodian-as-principal, to manage their hedging programme, will lead them to achieve the best possible price on which to execute business.
“By instituting control procedures rooted in independence and objective measurement, investors can expect lower transaction costs in FX and better returns to their portfolios.”