FMP 3: Fund Management

Open-end funds sort into three families, with a fourth for anything that straddles them. Money market funds hold fixed-income instruments maturing in under one year, commercial paper being the standard example. Bond funds hold fixed-income securities running longer than a year. Equity funds are the most popular of the three. A fund holding more than one type of security is a hybrid fund, or multi-asset fund.

Money market funds and breaking the buck

For many savers a money market fund stands in for a bank savings account. The return is usually higher than on bank deposits, and some funds let clients write checks on their holdings. One difference is decisive: there is no deposit insurance.

Most money market funds in the United States hold the NAV constant at USD 1, with each day’s gains returned to the investor rather than accumulated in the share price. A negative return pushes the NAV below USD 1, and the phrase for it is breaking the buck. The event is rare. It happened to the Reserve Primary Fund in September 2008, after it was left holding commercial paper issued by Lehman Brothers and Lehman Brothers declared bankruptcy. The government stepped in with a guaranty to avoid a panic. Regulations from the Securities and Exchange Commission have since imposed a floating NAV requirement on some money market funds.

Once each trading day, at 4 p.m., an open-end fund values everything it holds. Divide that total by the share count and the result is net asset value per share.

For an open-end fund the NAV is not merely a reported statistic, it is the transaction price. All sales and purchases of shares happen at the 4 p.m. NAV struck after the order is placed. Someone who telephones a broker at 10 a.m. cannot know what price the trade will get, since nothing is executed before 4 p.m.

How the investor is taxed

Some savings vehicles carry a tax benefit, and money placed to provide an annuity or a pension is tax deferred in many jurisdictions. A mutual fund carries no such benefit. Tax authorities look through the vehicle and treat the holder as the direct owner of what it holds, so gains realised inside the portfolio are taxed in the year they arise. Shares bought by the fund at USD 70 and sold at USD 90 hand the investor a capital gain of USD 20, and tax is due on it.

An index fund sets out to track a chosen benchmark, the FTSE 100 or the S&P 500 for instance, and is judged on how closely it managed it. Tracking error is the measurement. The popular version is the root mean squared error: square each yearly gap between fund and benchmark, average those squares, and take the square root.

What the fund charges

The recurring charge is the expense ratio, levied yearly as a percentage of assets under management. An investor may also face a front-end load, paid on the way in, or a back-end load, paid on the way out. Fees are relatively low in the United States and Australia and relatively high in Canada and most European countries, and they differ far more by style than by country. Active managers usually take between 1% and 2% a year, while a few United States index funds charge nothing at all, an offer Fidelity Investments opened in 2018. Part of the underperformance measured above is the management fee working through the returns.

During 2008, index equity mutual funds were estimated at 14% of all equity mutual fund assets. By 2018 the figure for index mutual funds had reached 29% of assets across all mutual funds.

A closed-end fund keeps the same number of shares through time. In essence it is a regular company whose business is investing in other companies, so buying into it is like buying a share in any other company. Price changes balance investors who want to sell against investors who want to buy.

That difference produces a price puzzle. Open-end fund shares change hands at NAV, whereas closed-end fund shares typically trade below it. A fund whose NAV is USD 32 per share might see its shares quoted at USD 30. In theory an arbitrageur could buy every share at the market price and sell the underlying assets, but the attempt bids the share price up and the profit disappears.

The management fee explanation

Stephen Ross put the discount down to fees. Buying a fund share buys a return stream already shaved by the manager’s charge, while the underlying assets bear no such charge, and a smaller stream discounted at one rate must be worth less.

Closed-end funds are far less popular than open-end funds, but they hold two advantages. Their shares change hands whenever the exchange is open rather than once at 4 p.m., and they can be shorted, which means borrowing stock from another holder and selling it in the hope of a cheaper repurchase later. They also need not hold liquid assets against redemptions, because closed-end fund investors trade with each other.

An exchange-traded fund borrows from both of the structures above and adds a mechanism that pins its price to NAV. A 2008 survey of investment professionals ranked ETFs as the most innovative product to reach investing in the previous two decades. Trading began in the United States during 1993 and in Europe during 1999. The best known is the SPDR S&P 500 ETF, listed under the symbol SPY, which follows the S&P 500.

Creation happens when an institutional investor hands over a block of shares and takes ETF shares in return, after which those shares trade on an exchange like any other company’s. The privilege given to institutional investors is what matters: they may surrender ETF shares for their slice of the underlying portfolio, or run the trade the other way. Anything delivered in must match the composition of what the ETF already owns.

Because the trade runs both ways, price and net asset value do not drift apart significantly or for long. Many ETFs track an index, but an ETF can be actively managed, and the mechanism works only if institutional investors know what to deliver, so an ETF discloses its assets twice a day. Mutual funds disclose far less often.

Source: chapter discussion of the three fund types.

ETFs are gaining in popularity, and by 2018 their total net assets in the United States were about USD 3.4 trillion.

Mutual funds and ETFs in the United States are heavily regulated by the Securities and Exchange Commission. Prospective investors must receive complete and accurate financial information, and rules exist to prevent conflicts of interest and fraud. Despite those safeguards, four abuses recur.

Late trading

Trade instructions are supposed to reach a broker before 4 p.m., because that is the price they will receive. For administrative reasons the broker may not pass those orders to the fund until well after 4 p.m., and the delay creates the opening. If news breaks shortly after 4 p.m., a trader can call the broker to cancel an order due at the 4 p.m. price, or enter a new one at a price already known to be stale. Some brokers in the United States have dishonestly accepted such orders. Late trading is not permitted by the SEC, and several prosecutions have followed, producing large fines and costing the employees involved their jobs.

Market timing

Not all the assets of a fund trade actively, and prices that have not moved for hours are stale, meaning they no longer reflect recent information. Securities traded on overseas markets can be stale because of time zone differences. Suppose the time is 3:45 p.m. and markets have been rising for several hours. Some prices feeding the NAV predate the rise, so the shares are probably worth a shade more than the NAV about to be struck, and buying at it is attractive. Where markets have been sliding instead, the attractive side is selling into that NAV. Market timing is not illegal, but the trades must be quite large to be worthwhile, and trades of that size whipsaw the fund up and down. The fund may then keep additional cash to meet redemptions, at a cost to every other investor. Regulators are concerned when special trading privileges are offered to market timers.

Front running

A trader working for a fund learns that the fund is about to execute an order large enough to move the market. Suppose the fund is about to purchase 1 million shares in one stock. The trader takes 10,000 for a personal account first and lets the fund’s own order lift the price. Front running is illegal in fund management.

Directed brokerage

The last practice involves no trading edge. A mutual fund and a brokerage house reach an unwritten agreement: the brokerage house recommends the fund to its clients, and the fund uses it for trades. The arrangement buys distribution with money belonging to the fund’s investors, and regulators frown on it.

Hedge funds are a form of alternative investment and are subject to much lighter regulation than mutual funds and ETFs. Mutual funds and ETFs serve small investors. A hedge fund takes money only in large amounts, and only from wealthy private individuals or from institutions, and the differences following from that run through the whole arrangement.

Source: comparison of hedge fund and mutual fund terms as set out in the chapter.

A schedule described as 2 plus 20% charges investors 2% of the value of their investment per year and adds 20% of the profits whenever those net profits are positive. Investors wanting a spread of hedge funds can buy a fund of funds, which picks a portfolio and charges its own fee on top. Fund of funds used to charge as much as 1 plus 10%, though their fees are now usually much less.

Where the name came from

The label comes from the long-short strategies the earliest of these funds ran, buying stocks expected to provide good returns and shorting those expected to provide poor ones. Carol Loomis coined the term in 1966 in an article about A.W. Jones & Co., the first hedge fund. The description has drifted from the practice, since many hedge fund strategies involve little or no hedging.

Jim Simons, a former math professor, founded Renaissance Technologies in 1982. Its flagship Medallion fund returned an average of 35% per year over a 20-year period, a figure that includes a 98.2% return in 2008, the year the S&P 500 lost 38.5%. Performance of that order changes what a manager can charge, and Renaissance reportedly took as much as 5 plus 44%. Estimates put the wealth of Jim Simons at USD 21.5 billion in 2019. The other names cited most often are Bridgewater under Ray Dalio, Soros under George Soros, and Citadel under Ken Griffin, whose incentive fees made all three founders very rich.

Two conventions must be fixed first, since the same schedule gives different numbers under different ones. Here the management fee rests on assets held when the year opens, and the incentive fee is struck once that management fee has been taken out. Some hedge funds base the management fee on end-of-year assets and calculate the incentive fee before it is deducted, which is more aggressive.

Write A for the assets under management at the start of the year and R for the return during the year. Under a 2 plus 20% schedule the incentive fee is as follows.

Read the expression as a call option held by the manager, written on the dollar return, whose strike sits at 2% of assets managed. Below the strike the fee is zero however badly the fund does, and above it the fee rises at 20 cents per extra dollar earned, so the hedge fund has an upside but no downside.

What the investor keeps

Turning the fee around gives the investor’s return as a function of R. Above 0.02 the investor keeps 0.8 multiplied by the quantity R less 0.02; below 0.02 no incentive fee arises and the investor keeps R less 0.02. Under the aggressive convention the investor keeps 0.8R less 0.02 multiplied by the quantity 1 plus R when R exceeds zero, and R less 0.02 multiplied by the same quantity when R is below zero.

A fee that runs in one direction only does not escape investors, and funds routinely soften their terms to answer the objection. Three devices do most of that work.

The hurdle rate

A hurdle rate sets a level of return that must be exceeded before any incentive fee applies. It moves the strike price of the embedded option higher, so the manager is paid for performance rather than for a rising market.

The high-water mark

Under a high-water mark clause the incentive fee applies only where an investor’s cumulative profits are positive, so losses have to be recovered before the fee restarts. If USD 100 million is invested and the hedge fund loses USD 10 million, it must make USD 10 million before incentive fees kick in. Such clauses often carry a proportional adjustment, meaning the mark covers only money left in place. Should the investor pull out half of what remains after the loss, the hedge fund need make only USD 5 million before incentive fees apply again.

The clawback

A clawback clause allows incentive fees already paid by the investor to be used to offset future losses. A manager paid on a strong year who then gives the gains back does not keep the whole of the earlier payment, which is the closest these devices come to a genuine downside.

None of the three fully solves the problem, because a manager sitting on substantial losses has a way out. Closing the fund and starting a new one leaves the old high-water mark and the old clawback behind, and the option embedded in the new fund’s fees starts again at the money.

An option gains value from volatility, and the incentive fee is an option, so the manager gains from volatility too. Take a hedge fund with USD 100 million of investors’ funds on fees of 2 plus 20%, and put six strategies in front of it.

Strategy A1 is a safe investment producing a profit of USD 3 million with certainty, an expected return of 3%. Strategy B1 has a 50% chance of a profit of USD 5 million and a 50% chance of nothing, an expected return of 2.5%. Strategy C1 has a 50% chance of a profit of USD 10 million against a 50% chance of a loss of USD 10 million, an expected return of 0%. As investments, B1 and C1 make no sense, since the expected return falls as risk climbs.

A second set rewards risk properly, so the criticism cannot rest on a rigged choice. Strategy A2 is again a safe profit of USD 3 million. Strategy B2 has a 50% chance of a profit of USD 10 million and a 50% chance of nothing, an expected return of 5%. Strategy C2 has a 50% chance of a profit of USD 30 million against a 50% chance of a loss of USD 12 million, an expected return of 9%.

Source: chapter Tables 3.1 to 3.4, combined.

The management fee is USD 2 million in every case, being 2% of USD 100 million, and the incentive fee is where the strategies part company. Strategy A1 leaves a profit of USD 1 million after the management fee, so the incentive fee is USD 0.2 million. Strategy B1 pays nothing half the time and 20% of USD 3 million the other half, while C1 pays nothing half the time and 20% of USD 8 million the other half. The worst investment carries the largest expected fee, and every step that improves the manager’s position makes the investor worse off, ending in an expected loss of 2.8%.

The same pull when the trade-off is fair

The second set removes the perverse trade-off and the pull survives. Strategy A2 matches A1, so investors again earn 0.8%. Under strategy B2 they earn 2.2%, being 5% less 2.8% of fees, and under strategy C2 they earn 4.2%, being 9% less 4.8%. Everybody gains as risk rises this time, but the expected return of the hedge fund exceeds that of the investor in every row and is never negative.

Abuses of this kind do happen, and Amaranth is the standard illustration. Brian Hunter, one of the firm’s star traders, built very large leveraged positions in natural gas futures during 2006, betting that winter prices would climb against summer prices. The same trade had worked a year earlier, when hurricanes adversely affected natural gas supplies, and his bonus for that year was reported at roughly USD 100 million. In 2006 the strategy lost about two thirds of the USD 9 billion invested with Amaranth. The fund was wound down. Hunter kept the bonus.

Long-short equity

The original hedge funds purchased stocks considered underpriced and shorted those considered overpriced. Done properly the strategy strips the market out, leaving performance that depends only on the manager’s ability to pick winners and losers. Two conditions have to hold: the value of the shares shorted must equal the value of those bought, and both portfolios must have the same sensitivity to market movements.

Suppose Ford and General Motors are judged equally sensitive to moves in the S&P 500, with Ford undervalued and General Motors overvalued. A long-short strategy could then involve buying USD 100,000 of Ford stock and selling USD 100,000 of General Motors stock, so that whatever the index does affects both legs alike.

Dedicated short

At any given time there are probably as many overvalued shares as undervalued ones, and a dedicated short fund devotes its attention to picking the overvalued ones, looking for companies experiencing difficulties the market has not recognised. Nothing hedges the market exposure, so dedicated short strategies perform badly during bull markets.

Distressed securities

Some hedge funds specialise in distressed securities, trading the debt of companies in serious difficulty. Knowing how bankruptcy works lets them spot cases where a big position in the debt turns a reorganisation proposal to their advantage.

Merger arbitrage

An announcement that one company will buy another leaves genuine doubt about completion. The target’s share price usually rises on the news but stops short of the price being offered. A fund putting the chance of success at 80%, and expecting the final price to beat the current one, has reason to buy the target. Where the terms are a share-for-share exchange that looks likely to improve, the fund buys the target and shorts the acquirer at a ratio set by the offer on the table. Inside information plays no part in any of this, and it is illegal: Ivan Boesky served three years in prison for insider trading.

Convertible arbitrage

A convertible bond can be converted at a future time into a predetermined number of the issuing company’s shares. Convertible arbitrage funds run sophisticated models over these bonds and trade whenever the market price parts from the model price, hedging their exposure to the issuer’s share price, to credit spreads and to interest rates. Profit arrives if the market price converges to the model price.

Fixed-income arbitrage

Among otherwise similar bonds, some look relatively expensive and some relatively cheap. A fixed-income arbitrage manager buys the cheap ones, shorts the expensive ones, and waits for the relationship to normalise. The mispricings are small, so the strategy typically uses a lot of leverage to be worthwhile, and the leverage is what makes it dangerous.

Emerging markets

Funds working in emerging markets build knowledge of thinly followed equity securities in developing countries, buying locally what looks undervalued and shorting what looks overvalued. Another route is the American Depositary Receipt, a certificate backed by shares of a foreign company and traded on a United States exchange, since any gap between the receipt price and the local price is an arbitrage. Emerging market sovereign debt is open to these funds too, and it is fraught with risk: Russia, Argentina, Brazil and Venezuela have each defaulted multiple times.

Global macro

Global macro funds use macroeconomic analysis to find markets that are not in equilibrium, with models built on exchange rates, interest rates, balances of payments and inflation rates. Results are sometimes spectacular. George Soros ran the Quantum Fund, which took a profit of USD 1 billion during 1992 from a bet that the British pound was overvalued. Most such trades end less dramatically, and an economy can sit out of equilibrium for years at a stretch.

Managed futures

Managed futures strategies attempt to predict future commodity prices and take positions that pay off if the predictions are correct. Several models are used, and the rules get back-tested, which means checking what they would have delivered had they been running in earlier years. That test cannot tell a rule grounded in a fundamental understanding of the market apart from one that was merely lucky, and only the first sort keeps working. The defence is out-of-sample testing, keeping the data used to develop a strategy separate from the data used to test it.

The risks running underneath

These strategies share a small set of exposures. Leverage magnifies losses as readily as gains. Convergence trades rely on a model, so model error is a live exposure, as is the chance that convergence takes longer than the fund can wait. Liquidity can vanish when a position needs closing, and a market neutral strategy is neutral only while its assumptions survive.

The prime broker of a hedge fund is the bank handling its trades and supplying its borrowing. Short positions are common, and the prime broker arranges those as well, frequently adding risk management and hedging services. Stress tests run by the bank over the portfolio set the size of the credit line, and the fund posts its securities there as collateral.

Hedge funds face very little regulation, yet the prime broker constrains what they can do. In an adverse economic environment, such as the one experienced during the 2007-2008 crisis, it may cut the borrowing limit and compel the fund to unwind positions.

Short-term losses on a strategy that would eventually work

A strategy can be close to certain of paying off over a long horizon and still put the fund through heavy losses on the way, and those losses trigger calls for additional collateral. A fund therefore has to weigh how much interim loss its prime broker will keep financing. Long Term Capital Management took positions that could reasonably have been expected to be profitable if held for several years. Russia defaulted on its debt in 1998, the interim losses were huge, and the collateral calls that followed destroyed the fund.

The risk that ran the other way

Prior to 2008 it was considered that nearly all the risk in the prime broker relationship sat with the prime broker. The Lehman default corrected that. Funds that had used Lehman Brothers in that role discovered they could not reach the securities posted there as collateral, and the market understood from that point that both sides carry risk.

What an investor wants to know is whether hiring a professional to run the money pays. The evidence does not support a confident yes. Excellent returns have certainly come out of some mutual funds and some hedge funds, but every solicitation carries small print that is entirely true: past performance is no guarantee of future results.

What the research on mutual funds found

Do actively managed mutual funds beat the market on average, and does a fund that beats it this year stand a good chance of repeating that next year? Michael Jensen looked at both questions in the 1960s, and on both counts the answer looks like no, a result later researchers working with more recent data have confirmed.

The first finding is unsurprising once the arithmetic is laid out. Before expenses, everybody’s returns together are the market’s return. Over 30% of all United States corporate equity sits with mutual funds and ETFs, so any systematic gain by that block would require a matching systematic shortfall for everyone else. Once expenses come out, the actively managed funds fall short on average.

The second finding concerns persistence, the tendency of good performance to repeat. Of the mutual funds that beat the market in a year, only 50% managed it again the year after. A two year run left roughly a 50% chance in the third year, and runs of three, four, five and six years behaved the same way.

Advertised track records look like a contradiction until the selection is accounted for. The fund in the advertisement is one of many offered by the same house, and if beating the market in any year is a coin flip, the chance of doing it four years running is 1/16. A company with 16 different funds has a good chance that one managed exactly that, and that is the fund advertised. This research pushed money toward index funds, which are termed passive investments and charge lower fees.

Why hedge fund numbers are harder to trust

Assessing hedge fund performance is harder, and the difficulty lies in the data rather than the analysis. Organisations do gather returns and publish statistics by strategy type, but taking part is to some extent voluntary and plenty of funds stay out. The ones least willing to file are those sitting on losses and those already shut, which pushes average reported returns upward. That is also why statistics from one organisation do not always agree with those from another.

Source: chapter Table 3.5, BarclayHedge index and S&P 500 returns including dividends.

Hedge funds performed quite well against the S&P 500 before 2008, and the record since is poor. During 2008, a watershed for equity markets, hedge funds lost money on average and still beat the index by a wide margin. From 2009 to 2018 they underperformed it in every year. One explanation is that these funds lag when equities rally and hold up when equities fall, since most of the strategies are not built to ride market trends. The long-short strategy is the clearest case, being built to attenuate or eliminate the effects of market moves.

Set against that record, the ability of hedge funds to attract investors is striking. The HFR Global Hedge Fund Industry Report put assets under management at a record USD 3.245 trillion in mid-2019. Some funds have created a great deal of wealth and others have failed badly after several good years.