Quants clone private equity: pale imitation or real deal?

Theory says replication can work, but investors are reluctant to give up private equity’s smoothed returns

Ten years ago Bain Capital set out to analyse the private equity sector’s so-called secret sauce. The $100 billion investor put to work a team of analysts to gather data on deals by the top 25 private equity shops, itself included, hunting for patterns in how they made money. The analysts got a surprise.

The study, which looked at buyout deals across a period of 30 years, showed the “vast majority” of the profits in the industry had been generated by the cheapest deals, says Daniel Rasmussen, one of the analysts on the project. Private equity’s big idea, it seemed, was not much cleverer than buying low and selling high.

That got Rasmussen thinking about whether a similar bargain-hunting strategy might work just as well – or even better – in public markets.

Today Rasmussen runs Verdad, the hedge fund he founded in 2014 with that precise goal. Verdad, which runs “close to $200 million” in assets, is one of a small group of firms that have launched or are developing products to replicate private equity’s risk and returns using listed securities.

The replicators are, in general, quant houses and systematic investors. Their pitch is to provide an easily accessible and cheaper way for money managers to reap private equity-like performance, in an investible fund format. The traditional method of accessing private equity – via a capital allocation that may sit idle waiting for the right target – is seen by some as unpredictable and pricey.

Man Numeric launched a replication product in 2018. Neuberger Berman’s private equity division is trialling replication strategies in dummy portfolios. Other firms active in this area include boutique investor SummerHaven Index Management, which has created two exchange-traded funds that replicate private equity.

Succeed and the quants could snatch a chunk of the riches invested in private equity – $3.7 trillion in the past five years alone – much as passive funds stole business away from discretionary firms in mainstream investing, led by Vanguard’s pioneering index funds.

“When Jack Bogle launched Vanguard the take-up was very low for the S&P 500 index. No one wanted it,” says Julian Klymochko, founder of Accelerate Financial Technologies, a firm that offers private equity replication ETFs alongside other funds. “Today nothing beats it in terms of what investors want.”

The comparison may seem fanciful to detractors of private equity replication, and the meagre returns of the strategy so far urge investor caution. SummerHaven’s main strategy is down by around 5% since launch nearly two years ago, although year-to-date returns are up by 10%. Accelerate’s ETF is down 15% since inception in May.

These performance figures suggest the replicators have a long way to go. Private equity managers insist replication won’t work for a host of reasons; replicators themselves disagree on how exactly to do it; and institutional investors are loth to give up the low volatility of traditional private equity funds even though critics decry it as an accounting trick.

Yet there are grounds to think the quants can prosper. Chiefly, the cost savings will be hard for allocators to ignore. Fees for replication products are around a quarter to a fifth of those for conventional funds, based on the firms Risk.net spoke to.

In the Vanguard

The would-be Bogles of private equity broadly have formed three strategy blueprints.

A first group is focused on buying cheap micro- or small-cap stocks – a strategy that assumes historical private equity returns have come, as Rasmussen believes, from a play on the well-known value and size factors seen also in public markets. This group is not trying to faithfully reproduce the activity of private equity funds, rather adapt the sector’s principles and apply them to public markets with a factor investing twist.

Other firms believe that private equity’s success lies in its shrewd sector bets. These replicators are seeking to track the allocations of private equity funds and estimate what funds are likely to be doing using models based on past behaviour. In mid-2019, for example, technology was the most favoured sector, energy and utilities the least.

A third set of firms mimic the deal-by-deal thinking of private equity managers, capturing in algorithms the qualities that private equity managers look for in the companies they buy: fundamental indicators such as debt loads, cashflows, sales numbers. Critically, those preferences change depending on the macro environment, the quants say.

The rationale for replication appears strong – if past research is anything to go by. A 2015 study by Erik Stafford, Harvard Business School finance professor and an adviser to SummerHaven, found private equity buyouts tend to target smaller, profitable companies with low enterprise value to Ebitda ratios, and with a preference for share buybacks over new equity.

Stafford created unlevered long-only portfolios of listed public companies that fitted these characteristics, and modelled them over time. “And lo and behold,” says Kurt Nelson, head of client solutions at SummerHaven, “the long-term gross returns of these securities looked a lot like the net-of-fee returns of actual private equity.”

By buying stocks in public companies that have the properties of buyout targets, an investor could harvest similar returns to private equity, Stafford argued. “The results indicate that sophisticated institutional investors appear to significantly overpay for the active portfolio management services associated with private equity investments,” he wrote.

SummerHaven’s two ETFs are benchmarked to a pair of public indexes which replicate private equity funds. Verdad and Accelerate Financial Technologies employ strategies along similar lines (see box: Private equity copycats).

Chicago-based DSC Quantitative Group has created a benchmark index by tracking the returns of private equity-backed companies using data captured by Refinitiv and seeks to match the industry’s sector bets.

Man Numeric has built a model to mimic private equity dealmakers’ proclivities. The firm has drawn on data from buyouts and leveraged loans as well research from academics such as Paul Gompers at Harvard Business School into how private equity managers operate. The model also includes a qualitative element, using feedback from industry participants and the intuition of the firm’s own managers.

 

Then and now

It’s early days and replication firms are yet to hear the distant thunder of stampeding investors. Across the funds Risk.net spoke to for this article total assets under management would be counted in hundreds of millions rather than billions.

Backtested performance looks persuasive. Live performance numbers, where they are available, present a less impressive picture right now.

Verdad’s strategy would have delivered an average annual return of more than 24% from 1965 to 2013, the firm says. Annualised returns for DSC Quant Group’s index calculated from 1996 – with live performance from 2014 – beat the Cambridge Associates private equity returns index by nearly three percentage points a year (see graph).

Yet Klymochko says “typically, investors want to buy what’s been going up recently. Leveraged small cap value has not been the best strategy lately. It’s been quite challenging as the S&P 500 has performed astonishingly well on the back of large cap growth”.

For their part, private equity practitioners are quick to point out other difficulties facing the new funds.

The data that replicators rely on is sparse and lagged. Fewer than a thousand public-to-private transactions took place in the 10 years following the financial crisis, making it hard to be certain of patterns in the types of companies private equity firms like to buy. Private equity firms also release cashflows only quarterly with a three- to six-month delay in most cases.

Private equity managers say their methods have changed, undermining the validity of models built on historical data. A senior executive at one private equity firm estimates half its returns today derive from steps to boost sales and profits at companies in its portfolio, much different from the past when “private equity investors made money by identifying the right sectors, buying cheap companies and doing some financial engineering – in other words, loading them with leverage”.

Smooth operators

The single biggest impediment to the take-up of replication, though, concerns a feature of private equity the imitators have no way to match. Private investments are not marked to market but to internal models, meaning cashflows are artificially smooth.

The effect is twofold. First, the cashflows are less useful for calibrating replicator models that re-create private equity returns. Quants can try to de-smooth the data using a statistical approach that involves estimating likely higher-frequency numbers, but the technique is far from exact.

Second, marking by internal valuation means private equity funds are able to report much lower volatility than public equities. The effects can be dramatic. In Q4 2018 the S&P 500 slid more than 13.5%, for example, while the Cambridge Associates US private equity index, which tracks valuations as reported by private equity firms, was down just 1.7%.

Investment mandates may prevent some money managers from sinking capital into funds that exhibit high volatility, giving private equity firms a built-in advantage.

Replicators are damning about this practice. “Private equity is built on a statistical fiction, but it’s a powerful one,” says Andrew Beer, managing partner at Dynamic Beta Investments. “It’s a luxury for allocators to not have to look at quarterly marks since it reduces pressure on investment teams when equities decline.” AQR researchers have suggested that smoothing is one reason institutional investors might be willing to pay private equity’s higher fees.

All of which creates a disincentive even for curious allocators to give replication a try.

“You won’t know for a decade or longer, through a cycle or two, whether the replication approach was truly effective over any shorter timeframe, given specific choices about how to replicate which private equity funds,” says Eric Wetlaufer, former head of public market investments at the Canada Pension Plan Investment Board. “In the meantime, if private equity is making money, why change? A lot of people don’t want to be the first to do it.”

Private equity is built on a statistical fiction, but it’s a powerful one

Andrew Beer, Dynamic Beta Investments

Another barrier to persuading sceptical investors is the lack of single, cohesive argument for replication. Advocates of the strategy are a loose coalition of academics and practitioners, each with their own idea of an effective strategy.

Arthur Bushonville, DSC’s founder and chief executive, says a simple factor-based approach to replicating private equity falls down because factor loadings change too much in different market environments. DSC thinks sector tilts are where private equity’s profits are anchored.

Ludovic Phalippou, a finance professor specialising in private equity at the University of Oxford Said Business School, has a different take. His research suggests private equity, infrastructure or venture capital funds are at least partly replicable using well-known public factors such as size and value.

At SummerHaven, Nelson says the changing sector weights of traditional private equity are simply a by-product of funds’ asset selection process.

Rasmussen thinks close mimicry of private equity funds’ contemporaneous activities is an error.

“We should have no interest in doing what private equity is doing today,” he says, in reference to private equity firms buying companies at inflated multiples. “What private equity is doing today is procyclical and highly levered and risky. What you want to do is what private equity used to do in the 1980s and 1990s, when they were really generating massive outperformance.”

Coming around

Something that unites many replicators is a belief that private equity is due a stumble. AQR has raised the question of whether elevated private equity deal multiples will prevent the sector matching past performance in future. Average multiples in 2003 were about seven-times Ebitda, a common measure of earnings. Since the global financial crisis they have climbed to almost 11-times.

Returns in recent years have also markedly dipped. Consultancy Bain & Company says average gains for deals done by private equity firms since the financial crisis have drifted lower, to about two-times invested capital compared with three-times in the early 2000s.

At the same time, replication need not be perfect to be useful. It could serve to complement traditional investments. Anthony Tutrone, global head of NB Alternatives, a subsidiary of Neuberger Berman, says replication can offer better than cash returns for investors who are wary of parking their capital until a private equity fund accepts it. Investors in private equity funds typically commit capital up front but then must wait until that capital is “called” by the fund and invested. That can leave them undershooting their desired allocation to the sector for periods.

There are signs, also, that investors are changing their views.

Jean-François L’Her, senior adviser for asset allocation at the Abu Dhabi Investment Authority, alongside a team from the Canada Pension Plan Investment Board, published research in 2016 showing much as Stafford did that buyout funds track a levered size- and sector-adjusted index.

Notably, the Abu Dhabi authority and CPPIB are among the top three allocators to private equity globally with approaching $100 billion in the sector between them, Preqin data shows.

Other allocators, meanwhile, are looking more generally to make savings. The California Public Employees’ Retirement System reported average total fees for private equity over 20 years of 7% a year. The pension fund is today looking to take closer control of its private equity investments to reduce costs.

“There’s been a lot of talk in the media about not lining the pockets of private equity billionaires for unspectacular returns,” observes Wetlaufer of CPPIB.

And, as Wetlaufer sums up, cheapness could prove to be replication’s killer feature. “Replication is not a trivial exercise,” he says. “But none of this is rocket science either.

“The thing is, it’s possible – and you don’t have to pay two and twenty to do it.”

Private equity copycats

How do replicators approach their task?

SummerHaven’s general buyout strategy index selects 300 to 350 stocks ranging from micro-cap to mid-cap that have low enterprise value to Ebitda ratios, solid profitability, and exhibit signs of confident management, as evidenced by recent share buybacks rather than equity issuance.

Verdad’s strategy filters US small-caps – generally under $1 billion in market cap – to find stocks with average valuations below seven-times Ebitda that have the potential to benefit from paying down debt, so elevating their stock value. The firm’s portfolio managers exercise discretion on investments after stocks are filtered.

Accelerate elected in its zero-fee long-short ETFs to filter North American stocks by value, size and leverage, taking relatively concentrated positions, with fewer than 50 stocks in each of the long and short portfolios.

DSC takes data on private equity holdings from Refinitiv to set a basis for sector weights in its portfolio, then uses more than 20 models based on mainly macroeconomic inputs to develop a consensus forecast on how much leverage to take on, to match the market sensitivity of the benchmark.

The firm re-allocates that leverage sector-by-sector, using its models of what private equity firms are likely to be doing in the current environment – tilting the portfolio towards some sectors and away from others. The final step generates three to four times the value of the others, the firm says.

The firm’s investible index tracked the Refinitiv benchmark index – in other words, the estimated monthly gross returns of private equity investments – within 40 basis points over that period.

Man Numeric’s model favours certain sectors and selects companies with features that buyout firms lean towards. According to the firm’s analysis, that tends to be small, cheap, profitable, growing companies that have the ability to take on debt.

The model prefers to steer clear of industries such as utilities or biotech, where regulation or other factors typically put private equity players off, says Gregory Bond, director of research at Numeric and one of the model’s creators. “Its behaviour is also influenced by the sectors that private equity capital is going into, like tech, healthcare, industrials, and consumer discretionary,” he adds. The portfolio trades about 200–300 stocks from the Russell 2500 Index.

Editing by Alex Krohn

• Tweet  
• Facebook  
• LinkedIn  
• Save this article
• Send to  
• Print this page  

• Topics
• Buy side
• Data
• Factor investing
• Indexes
• Institutional investors
• Investment
• Portfolio
• Private equity
• Quantitative analysis
• Volatility
• Asset management
• Investing
• Quant investing