All change as investors favour ETFs over futures

The use of futures to gain long equity index exposure is no longer the efficient play it once was, thanks to stubbornly high roll costs and the flight of arbitrageurs from the market. ETFs are being touted as a cheaper alternative, but is there more to the debate than meets the eye?

High roll costs and inefficient tracking spark shift to ETFs

A regime change is threatening to topple futures from their perch as the go-to instrument for institutional investors seeking long exposure to equity indexes. Higher costs, largely driven by regulatory constraints on banks' equity derivatives desks, have seen an increasing number of investors seek exposure using exchange-traded funds (ETFs).

Until now, the debate has been dominated by aggressive pitches from ETF providers keen to tout their wares as more cost-efficient. But the evidence is starting to stack up. In the first quarter of this year, one large European long-only institutional asset manager switched its holdings of Euro Stoxx 50 futures for a Euro Stoxx ETF; two UK firms exchanged their futures on Indian equities for an MSCI India ETF; and a UK hedge fund purchased a Topix ETF tracking Japanese equities and dispensed with their unfunded holdings.

In all three cases, firms cited the greater efficiency achievable in ETFs as the primary reason for the switch.

"We're in a new world order," says Emmanuel Dray, global head of equity flow derivatives at BNP Paribas in London. "I think we have to understand that we are living in a new era of regulatory constraints. The pre-2012 and post-2012 worlds are not the same, and I don't see the economics of futures coming back unless regulatory pressure eases."

The present cost differential between futures and ETFs puts weight behind such anecdotal claims. The total costs of holding a $100 million long S&P 500 exposure over a one-year holding period as of October 2014 stood at 26.6 basis points for E-mini S&P 500 futures and 13.2bp for one S&P 500 ETF, according to Bloomberg data compiled jointly by BNP Paribas and ETF provider BlackRock.

We’re in a new world order. The pre-2012 and post-2012 worlds are not the same, and I don’t see the economics of futures coming back unless regulatory pressure eases

The inflated cost of futures has been compounded for fully funded buyers by the vehicles' inability to cover the cost of funding through the yields generated on their margin and non-margin accounts. Such investors typically lock up a cash amount equal to the notional futures exposure in risk-free money market instruments - typically government bonds. Loose monetary policies around the globe have pushed yields on these instruments to historic lows, however.

"If you don't get Libor on your cash and you roll futures at a price above theoretical value, you lose ground compared to the gross return index you are tracking with futures," says Vincent Denoiseux, director, passive asset management at Deutsche Asset and Wealth Management (AWM).

Quantifying this cost is an imprecise science, since each investor will allocate their cash to different assets and different banks will offer varying rates on margin accounts. But research from Deutsche AWM suggests that a US-based investor allocated to three-month Treasury bills would have swallowed an underperformance hit of 49bp versus three-month Libor in the period since January 1999, and a whopping 211bp in the worst 12 months during that period.

Holding period

Not everyone agrees, however. Although momentum may appear to be behind ETFs, dealers argue that they are far from the clean sell providers make them out to be. Factors such as whether investors have a shortage or glut of cash and the length of holding period also have a major bearing on allocation decisions.

"If your holding period is short and there is an appropriate and liquid future available, investors should always consider using that future, since balance sheet costs or roll costs won't be such a factor," says one senior equity derivatives banker. "If you are holding for a longer period, especially over year-end when the cost of rolling futures typically peaks, and you're a real-money investor, then the balance really moves back towards ETFs – as long as you choose the right ETF."

This view is lent credence by investors who say futures remain the best bet for short-term or tactical allocations to particular equity markets.

"We use futures and options to get exposure to equity indexes," says Robert Andrén, head of equities at KPA, a Swedish pension fund. "We believe futures still provide the fastest and most efficient way to get in and out of different equity markets. Since our intention is not to hold the positions for a long time, we are not that worried about the increasing richness in the futures markets. We don't use ETFs because we don't have the cash and don't want the management fees."

Andrén says KPA's equity allocation is divided into a cash equity portfolio and a derivatives overlay portfolio. The cash equity is fully invested in stocks, while the equity derivatives overlay is for tactical and strategic asset allocation.

Dealers argue that ETF liquidity, which varies from region to region, also has a major bearing on costs relative to futures. "European futures still have a slight advantage over ETFs, whereas in the US you have very liquid books and can execute big sizes. In the US, if you want to trade a big block you can do it on screen; in Europe, many do so via a dealer, because of comparatively lower liquidity," says one London-based equity derivatives banker at another bank.

Tracking error

But ETF providers insist longer-term investors are being hit by a double whammy of higher futures roll costs and feeble returns on cash lodged as margin against those positions. When a dealer sells a future, they generally borrow the underlying stocks in order to hedge their position for the duration of the contract. This incurs a funding cost – generally the bank's overnight funding rate. Each bank will have its own funding rate, which can vary greatly.

The seller then subtracts the dividend revenue generated by this portfolio from the cost of the future, in order to compensate the buyer for the fact it is foregoing dividends by not holding the physical stock.

Finally, the seller factors in the implied repo rate to account for any gains or losses it could make by lending out the underlying stock it is using to hedge. When these three elements combine to form a price that is lower than the spot price, a futures contract shows so-called cheapness; when the price is higher, richness.

Long-term holders of futures also have to price in the roll costs charged for extending the duration of their exposure beyond the contract's standard three-month expiry. This is the difference in price realised from selling the initial contract and striking a new one on the same underlying. The difference varies depending on how close to expiry the investor elects to execute the roll, and from quarter to quarter as demand fluctuates.

Since 2012, futures on many leading indexes have exhibited persistent richness thanks to implied repo rates shifting – and remaining – negative, largely due to regulatory-driven constraints on banks' balance sheet capacity. Not only has this cost investors, it's also played havoc with banks' hedging strategies.

Such costs may be enough to push investors away from futures, ETF providers hope. "When they hold a future, financial institutions are exposed to a parameter that varies depending on regulatory pressures and speculation on the future cost of carry," says Guillaume de Martel, fund manager at ETF provider Lyxor in Paris. "Futures contain parameters that are random and volatile; on the ETF side, you have an easy-to-read parameter in the cost – represented by management fees – that is under downward pressure, because it's a competitive market," he says.

Inflated costs

Banks' ability to rent out their balance sheets to investors is also undergoing a drastic squeeze as regulatory requirements bite, says Deutsche AWM's Denoiseux.

"Balance sheet charges have increased with Basel III. Other reforms, such as the introduction of financial transaction taxes, plus the harmonisation of dividend withholding tax, are pushing up costs as well. Taken together this has increased costs for the market-makers, making it more expensive to provide synthetic equity to investors."

The top culprit is the Basel III leverage ratio, which requires global banks to hold a minimum of 3% capital against their leverage exposure. This rises to a maximum of 6% for US banks subject to the supplementary leverage ratio.

These requirements are yet to be formally implemented, but their influence is already being felt. In the five years from 2010, front-month Euro Stoxx futures have underperformed the Euro Stoxx 50 total return gross index by an average of 53bp, compared to an underperformance of 26bp over the 10-year period since 2004, according to Deutsche AWM research. Market observers say this doubling of the tracking error is a direct consequence of the squeeze on balance sheets imposed by the leverage ratio.

"Pre-crisis, banks were leveraged at 40 times their balance sheets; now, they are leveraged 10 times. A vast amount of leverage has come out of the system. This has caused a lot of friction where there used to be frictionless trading," says Michael John Lytle, London-based chief development officer at asset manager and ETF provider Source.

Another factor is the liquidity coverage ratio, which requires banks to hold liquid assets against short-term repo positions to protect them in the event of a liquidity crisis. This ratio has seen banks tighten the leash on their delta one desks, restricting their ability to use repo, and bumping up the total cost of funding for futures buyers.

In combination, these regulatory pressures have curbed dealers' appetite for short futures positions and caused the implied repo rate to dive into negative territory.

"It is a classic case of the unintended consequences of regulatory reform," argues Lytle. "Regulators think they can force up Tier I capital ratios and it will have no impact on the markets. But ultimately it hurts investors, in this case wealth clients, insurers and pension funds."

Historically, index arbitrage desks in banks and hedge funds would inject liquidity to the futures market when the implied repo rate turned negative, seeking a return from the difference between the rate priced into the future and the actual repo rate they could gain from lending out the basket of securities used to hedge the trade. But supply constraints have been exacerbated by the flight of arbitrageurs from the market.

Dray, among others, is convinced this is a structural change that will be hard to reverse. "I don't see implied repo rates springing back unless regulatory pressures ease. Index implied repo could also become less negative if the market turns bearish and fast money institutions are heavily selling futures," he says.

But others are optimistic that strengthened bank balance sheets will eventually smooth out regulatory effects. Anecdotal evidence from banks' equity derivatives desks suggests futures rolls on leading indexes in the current March/June quarter are trading at lower premiums. The picture also varies by index. Roll costs may have hit queasy altitudes for the FTSE 100 and S&P 500 but, for broader indexes such as the Russell 2000 and MSCI World, buyers can still be paid for going long.

Are ETFs the future?

While the funeral rites for index futures may be some way off, ETF providers are pushing buy-side clients to migrate to their fully funded index trackers as a cost-saving solution versus synthetic equity.

The cost of investing in ETFs is expressed in a total expense ratio, summing the management and administration fees charged by the provider. Investors also incur a bid-offer spread if they elect to sell out of their holdings. Providers claim these costs are now lower than the equivalent cost of index replication through futures.

Source cites the example of a fully funded investor seeking long Euro Stoxx 50 exposure. The effective cost for a futures investor over the period from March 2013 to December 2014 was 60bp. The equivalent cost for the Source Euro Stoxx 50 Ucits ETF over the same period meanwhile was 50bp, implying a 10bp performance spread in favour of Source.

The pattern is the same across the ETF universe, and holds across both physical and synthetic funds. Deutsche AWM's db X-trackers Dax Ucits ETF outperformed the corresponding index futures by 34bp a year in the period from November 2011 to November 2014.

The trend appears to be accelerating. Inflows to ETFs in recent years have increased liquidity and narrowed bid-offer spreads, while fierce competition among providers has seen management fees slashed. Denoiseux says Deutsche AWM has witnessed inflows of €650 million into its Dax db X-tracker over the past three months. In the year to date, Deutsche's platform has seen inflows of around €3.3 billion across its product range.

Events have also conspired to place ETFs in a favourable light. "One of the driving reasons behind the reallocation from futures to ETFs is quantitative easing. The world is cash rich, which means the need for leverage is less obvious than in the past. Some investors do not know how to use the cash they have, so a fully funded investment in an ETF makes sense," says Dray.

Withholding tax

One senior equity derivatives banker at a European bank argues that the current debate overlooks a key attribute that makes futures more attractive for buyers: the domicile of the investor, which can have a major impact on dividend-withholding tax treatment.

In the example of a Swiss investor buying a US-domiciled ETF holding German stocks, the investor would have received about 85% of the German stocks' dividends had he held them directly. However, by using the US-domiciled ETF, the Swiss investor could take a double hit by paying US withholding tax on the already-taxed German dividends.

"If your average market participant is eligible to receive a high percentage of dividends on stocks that they purchase, the futures basis will reflect a high percentage of those dividends – so futures get cheaper," says a senior equity derivatives banker at one bank.

The banker posits an example in which an index is worth $100, dividends are at 5%, and a future on the index expires a year from the strike date. Assuming zero interest rates, no withholding tax and zero balance sheet costs for the market-maker, a future over the index will cost $95.

The impact of dividend yield on the price is determined by the average tax treatment for all participants in the market. If a particular investor was subject to 20% withholding tax in this hypothetical market where the overwhelming majority of participants paid none, the future would still trade at $95, although fair value to that with-held investor would be $96.

On the ETF side, the relative value of investing varies depending on the respective domiciles of the investor, the ETF, and the underlying stock, as well as more obvious considerations such as management fees, tracking error, and the fund's stock lending policy.