Mariner’s CRO on avoiding predictable surprises

Buy-side risk survey: relative value specialist builds shocks into investment strategy

This is the fourth in a series of articles connected to our buy-side risk survey. Click here to read the rest of the series.

Friday the 13th of March was a real horror show for relative value hedge funds. After stocks crashed the previous day, investors began dumping bonds in a desperate dash for cash. Dealers struggled to absorb the selling. Spreads widened and the usual correlations between bonds and futures broke down, setting off a massive unwinding of basis trades. That morning, yields on 30-year US Treasuries fell seven basis points in less than 15 minutes – the financial equivalent of all hell breaking loose in the sleepy suburbs.

Dan Bradley, chief risk officer at Mariner Investment Group, was not surprised by what he saw that day. He is surprised others were.

“The structural illiquidity since 2008 has been well documented,” he says. “When you have forced sellers in a structurally illiquid environment, then you’ll see these violent price moves.”

In March, Bradley was head of risk at Corvid Peak, a hedge fund specialising in credit strategies that was previously called Tricadia Capital. Tricadia liquidated most of its assets in 2018 and 2019 following a management change and investor withdrawals. Corvid Peak manages Tricadia’s legacy investments and some new funds launched since 2019. In June, Bradley joined Mariner Investment Group, a hedge fund with $11 billion of investments, primarily in relative value strategies. Though he joined after the worst of the crisis, Bradley confirms Mariner was not among the forced sellers in March. “At no time was Mariner involuntarily liquidating to fulfil margin calls, or any other funding or financing obligations,” he says. Strains in liquidity and funding were factored in ahead of time, when sizing and hedging positions. As market conditions deteriorated in March, the firm responded in fairly routine fashion, by reducing risk, hedging and re-allocating capital among its three main strategies, across rates, mortgages and credit. Other firms had to take much more drastic measures. Relative value hedge funds were forced to offload tens of billions of dollars in government bonds. “Sharp changes in government bond prices, margin call and inability to rollover funding forced substantial sales of government bonds by these players (almost$90 billion during March) which in turn generated further falls in government bond markets,” Jon Cunliffe, deputy governor at the Bank of England for financial stability, said in a speech in June.

Bradley’s argument that investors could have seen the squeeze coming is grounded in changes to market-making capacity over the past decade. Dealers have smaller balance sheets and are constrained by an array of liquidity and risk-based capital ratios. The Volcker rule prevents them from taking proprietary risk. Inventories are kept to a minimum.

“The ability for banks and highly regulated institutions to absorb massive amounts of risk and provide liquidity to the market is not there, and it hasn’t been there since 2008,” says Bradley.

Investors got a taste of what this means for bond market liquidity during the European debt crisis in 2010 and the sell-offs in 2015 and 2016. Many were still wholly unprepared for what happened in March. “That liquidity risk has not been fully built into the fabric of the positions people are taking,” says Bradley.

Mariner prides itself on exactly this. Bradley says part of his job is to take the usual battery of risk measures – including value-at-risk – and share the results in a way that will inform positioning and portfolio management.

“At a high level, one of the things I’m focused on is reporting – how do we produce numbers that are actually meaningful?” says Bradley.  “Everyone calculates VAR, runs stress tests and computes different risk measures. We do all of that, but what we’re trying to do is find ways to make those numbers have more meaning in the context of our portfolios and the fundamental drivers of those trades.”

He gives the example of cash/futures basis trades, a staple of Mariner’s business. “Funding is really important in that trade. You have to consider the tenor of the trade and match that to the duration of your financing. Then, you have the fundamental drivers of that trade, which are linked to volatility and liquidity. All of that has to be incorporated into how you’re measuring the risk on that position,” says Bradley.

On top of that, the firm applies a range of extreme-but-unlikely shocks: “What if everything goes out of the window, and the basis blows out to two times the historical stress level? [The results] need to feed back into the investment process so it can inform our decisions on how we size positions and tail hedge the book,” he says.

Where’s the exit?

One of the challenges in measuring and mitigating liquidity risk specifically is that it can lie dormant for months or years at a stretch, as Risk.net’s recent buy-side risk management survey shows. Only 4% of respondents said liquidity was their most important source of risk in normal times – behind investment, operational and counterparty risk – while 38% ranked it second. In March, 35% said liquidity was their most important source of risk, just behind investment risk with 46%. A further 39% said liquidity was their second most important risk in March.

The on/off nature of the exposure makes it hard to quantify, Bradley notes: “When you’re evaluating an investment, a part of the return is liquidity premium. But it’s always very difficult to measure that – to say, 1% of the return I’m getting here is just pure illiquidity. It’s always one of the leftover pieces.”

According to the survey, the most popular measures of liquidity among investment firms are the time and cost to exit positions – cited by 74% and 60% of respondents, respectively. The catch is that these measures are usually calibrated to normal market conditions.

“When people talk about time to exit – which is the standard way the investment community has come to simplify it – you have to define the environment. And the environment people usually define is under normal market conditions,” says Bradley. “The problem is, when you get into these vacuums of liquidity, how do you measure it there? How do you measure the price at which you can get out when there is no market? That becomes a much more difficult exercise.”

This is the situation investors were facing in March. “We saw bond prices for credit instruments trade down 20 or 30 points, and it [was] a par piece of paper,” says Bradley. “Dealers were pricing bonds 10 points below the last transacted price. And if you went to the desk and tried to transact, it’s not going to be at that price – it’s going to be below that price, and that’s if you can transact at all.”

The market was effectively broken. “You get to a point in these crises where pricing signals are not necessarily real. They’re not transactable,” says Bradley. “That construct creates this velocity and causes price exacerbation.”

It was an unnerving experience for many traders and risk managers – the sort of thing that is not easily forgotten. Bradley expects liquidity risk to be taken more seriously as a result. “It is definitively going to be on the radar of risk managers and senior management. You’ll see it in how they’re defining their investment programmes, where they deploy leverage and where they choose not to deploy leverage,” he says. “It will become part of the investment decision, where people will ask ‘what investments are we making’ and then ‘how are we classifying liquidity’.”

To answer that second question, Bradley advocates stress testing. The practice is fairly routine in the industry – according to the survey, 76% of investors run stress tests using historical scenarios and 73% also use hypothetical scenarios. “That’s a positive thing,” says Bradley. “But I’m still surprised that’s not 100%. I would think everyone would be running stress tests and hypothetical scenarios as part of their risk practice.”

Check yourself

What he finds more worrying is that more than 40% of investors only run tests quarterly or annually. Fewer than a quarter run them daily. And more than 60% of firms that use hypothetical or custom scenarios only review them quarterly or annually.

“I was kind of shocked that the frequency [of stress testing] was anything other than daily. Portfolios are dynamic, markets are always moving, and inputs are always changing. I would expect people to be running these daily to see how they’re changing and evolving,” he says, adding: “Monthly is probably the borderline of how often you want to review the inputs.”

Mariner runs stress tests using hypothetical scenarios on a daily basis. The inputs to the scenarios are reviewed, updated and tweaked constantly, Bradley says.

The coronavirus crisis should prompt risk managers to step back and look at the larger picture when constructing scenarios. At most firms, this exercise tends to be inward looking, focusing primarily on financial and economic factors. This is understandable. Most financial stress events to date have been the product of endogenous factors. “It’s hard to think of one that was such an exogenous shock to the system,” says Bradley.

The roots of the 2008 financial crisis – loose underwriting, high leverage and disregard for counterparty risk – may have been hidden, but they were strengthening within the system for some time. Blow-ups in emerging markets are often the result of financial flows. “Money flows in, capital flows out – but it’s inside the ecosystem,” says Bradley.

“This was different in that it really came out of the blue,” says Bradley. “It shocked the system, but it had nothing to do with the system. It was a health crisis. It wasn’t part of the financial system but it totally disrupted the system. That’s one of the things that will change the way people think going forward.”

The entire episode revealed “the fragility of the system, whatever the source of stress”.

One consequence is that non-financial sources of risk such as climate change, geopolitics and technology may now feature more prominently in the hypothetical scenarios used in stress testing.

Risk managers will also need to account for the speed at which events unfolded in February and March. By comparison, the subprime crisis unfolded in slow motion, over a period from late 2006 to 2008. The first round of qualitative easing began in late 2008, with the second and third rounds coming in 2010 and 2012. The facilities to purchase troubled assets were not operational until the second quarter of 2009.

In March, central banks and governments responded to the emerging crisis with astounding speed and force. “This time, the monetary response was nearly instantaneous and actually went beyond what was done in 2008 – when balance sheet expansion was limited to mortgages and Treasuries – in terms of buying corporate bonds,” says Bradley. “The fiscal response was \$3 trillion right off the bat. The speed of that response was extraordinary.”

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe