Average US credit market valuations in October 2019 are nowhere near the levels seen in early 2016, let alone those in 2009. But according to Mill Hill founder and chief investment officer, David Meneret – who featured in The Hedge Fund Journal’s 2016 “Tomorrow’s Titans” report – certain relative value opportunities are at extremely extended levels. His proprietary measures of valuation dispersion, dubbed “credit basis”, are around the 99th percentile since Mill Hill’s inception and were only recently close to this point for a few days in 2016.
The term “basis” can refer to many things in credit markets – including cash versus CDS basis, or credit indices versus their constituents – but Meneret is not active in either of these areas, which tend to require substantial leverage to generate interesting returns. His concept of basis involves identifying two sets of instruments that are ultimately backed by the same, similar or reasonably overlapping assets or cashflows, but which trade at widely divergent valuations. Meneret thinks laterally among markets, trading a broadly market neutral portfolio across a wide range of instruments under four broad credit sub-asset classes: CLOs, financials, transportation & ABS, and MBS. In principle he is agnostic about whether the collateral is aircraft, company loans or residential or commercial property. The segmentation of other market participants who are unable or unwilling to cross over between credit sub-asset classes or trade all instruments within them – due to broker dealer capital constraints, investor comfort zones, narrow mandates and regulatory restrictions – results in blinkered thinking that can give rise to glaring valuation anomalies.
The big lesson of 2008 and even the 2011 crisis is do not be too levered when the markets get choppy.
David Meneret, Founder and CIO, Mill Hill
One axis of information asymmetry that segregates his investment universe is that between corporate and securitised markets. This creates inefficiencies partly because structured credit is a slower moving market, which typically takes weeks to respond to news that gets factored into other credit markets in a matter of hours, minutes or even seconds. The inefficiencies eventually get corrected partly because corporate and securitised markets are overlapping and interconnected. “For example, CLOs are comprised of leveraged loans, so if loans are cheap relative to CLOs, new CLOs will be created. If CLOs are trading below the value of their constituents, they can be called to crystallise the discount. Thus, there are mechanisms to engender re-convergence between the two related markets,” says Meneret. The strategy has actively traded CLOs, starting out with quite a high weighting in 2016, which was top sliced as they appreciated, compared to credit indices, into late 2018. The pullback in CLOs in 2019 then created attractive entry points for Mill Hill to rebuild the exposure.
In late 2019, Meneret has identified a divergence between leveraged loans and high yield bonds, which is partly caused by flows moving in opposite directions. Loans have seen months of net outflows while corporate bonds have recorded months of net inflows. Recently these technical flows have been driving the market, but Meneret, “expects that fundamentals, default and downgrade risks should be re-asserted at some stage. We are therefore long of selected CLOs against a short in corporate bond product, as the valuation gap between the two is at the highest levels since early 2009. CLO ‘BB’ tranche spreads have blown out, while investment grade CDS spreads have compressed to all time tights on one measure – currently at about 16.5 basis points of spread per turn of leverage, below the 21 basis points per turn of leverage level seen in June 2007. Yet there is a 52.7% (comparing BKLN and HYG indices) overlap between issuers of loans and issuers of high yield bonds, and the loans rank ahead of bonds in the capital structure.”
CLOs could now offer much more upside. Investors have switched from floating rate loans to fixed rate bonds partly as a way to play falling interest rates, but CLO yields may have touched their nadir. “Three-month LIBOR forwards for 2022 are now pretty close to the average LIBOR floor of 98 basis points on many loans,” Meneret explains. Moreover, the market is pricing in minimal reset or refinancing optionality in CLO mezz: many bonds trading at 85 or 90 could be reset or refinanced at par. This is possible partly because the ‘BB’ tranche is a small part of the capital structure, and the economics are largely driven by the much fatter AAA slice. The CLO market overall is weak, but the AAA segment remains very strong because European and Japanese banks are so desperate for yield. The compression of AAA spreads creates more headroom for upside on lower tranches,” says Meneret.
He continues that meanwhile, “investment grade credit has downside risks: the downgrade of Ford has already expanded the high yield market from circa $1.25trn to $1.33trn, and could soon be followed by Kraft Heinz, Anheuser Busch, ATT and GE. Investment grade is a big part of the short book”.
Meneret is selecting CLOs very carefully because spread widening is warranted on some of them. “Not all CLOs are equal,” he says. “Some high yield companies are deleveraging and getting upgraded, while some sectors such as retail face headwinds. Sectors such as technology and healthcare are superficially attractive, but upon lifting up the bonnet, it turns out that there is a lack of collateral which can imply a low recovery rate. Some CLOs may do well and others could perform poorly.”
He also declares that, “loans have underperformed this year, partly because issuer weighted default rates have ticked to 3.1%, the highest level since early 2016. There have been energy and retail defaults for some time and there is now elevated default risk in a supposedly defensive industry: healthcare, in firms such as Mallinckrodt Healthcare, Endo Healthcare and Envision Healthcare, which loaded up on loans. The healthcare sector has been adding 20% leverage a year for the past six years, but it is now vulnerable to the Senate cost-cutting bill. Recovery rates have also come down to around 60% on recent defaults, compared with earlier levels as high as 75%.”
Analysis of the loans market generates a thread of ideas for trades running through several other related markets that Mill Hill trades. “Firms with large exposure to leveraged loans, such as life insurance companies, may see some adverse impact from loan defaults,” he says, “and the extreme scenario of an insurer losing its investment grade credit rating can be hard to recover from. The last time this happened was Genworth five years ago, which is still languishing below investment grade. Some life insurers are more than 20 times levered. They are very levered, on multiple levels: their debt to EBITDA ratios, other liabilities, and investment portfolios. They have increased risk and duration, by reaching further out the curve for yield pickup and complexity premia from more esoteric and illiquid assets, investing in CLOs, CMBS and RMBS, often for small amounts of incremental yield. Life insurers are also shorts for us.”
“Loans have underperformed this year, partly because issuer weighted default rates have ticked to 3.1%, the highest level since early 2016,” says Meneret.
“Conglomerates with exposure to insurance, such as GE, which has underfunded long term care liabilities, could also be vulnerable. GE’s other problems include underfunded pension plans, and most importantly, exposure to aircraft engine making – which will see some fallout and cancellations from the Boeing 737-MAX accidents – and aircraft leasing, which requires a good credit rating. The aircraft leasing business essentially extracts a slim margin between borrowing and lending costs, and if borrowing costs spike after a company loses its investment grade rating, the business may not be viable. Aircraft leasing economics are also hugely sensitive to the assumed rate of depreciation,” says Meneret.
GE is a short leg and others in the space provide long ideas. “Secured aircraft bonds or enhanced equipment trust certificates (EETCs), and debt issued by aircraft leasing companies, should both gyrate with aircraft values,” Meneret asserts. “We look at the risk-adjusted fair value of long and short books, to work out the effective delta of each side of the trade. We then calibrate the ratio of the long to the short side so that we are agnostic about whether aircraft prices rise or fall, in order to isolate the valuation discrepancy between the instruments.”
Meneret is, as usual, selecting the longs very carefully. “We are cautious on the valuation of some aircraft backed securities, as values are high and lease rates are low. But certain niches in the market can also offer some value, in selected EETCs, which can also be a relatively secure investment. EETCs can carry on performing on principal and income payments even if an airline defaults on its lease payments, as airlines tend to be restructured rather than liquidated. Therefore, EETCs trading 20% below the value of liquid collateral, such as four-year old Boeing 737-800s (not 737-MAX), are good value. But the lack of demand for Boeing 737-MAX planes since the crashes makes other EETC values uncertain. Investors also need to distinguish between low-leverage EETCs backed by large airlines in developed countries, and highly levered ABS deals backed by aircraft leased to risky airlines, with lofty aircraft valuation assumptions.”
Mill Hill trades a wide variety of transport-related ABS and other instruments: secured airline bonds, aircraft ABS, auto corps, esoteric ABS, consumer ABS, and shipping corporations’ EETCs. They also look at related areas such as debt issued by aircraft lessors, railcar lessors, airlines, and aircraft manufacturers. The ABS world is a diverse one that includes a huge variety of underlying collateral, and Meneret has sometimes ventured into areas such as ABS backed by film royalty rights, attached to specific media companies, which in 2014 traded cheap relative to debt issued by those media companies. Today Meneret judges most esoteric ABS to be rather richly valued.
In the MBS bucket, Meneret has identified paper from a mortgage servicer and originator which has a double digit yield and paired it against 2-year CDS on a levered mortgage insurer, exposed to the first loss of risky mortgages, which costs over 70 times less – only 0.15% per annum. There is potential for either or both sides of this trade to move a long way, under benign or malign conditions for credit markets in general.
“Mortgage servicers have sold off as lower US rates increase prepayments and refinancing,” he explains, “and thereby shrink the asset pools they service, and the value of their mortgage servicing rights, but a deeper dive reveals that some servicers also have exposure to origination – which clearly benefits from more new mortgages being taken out. They have an intrinsic hedge from origination and prepayments have not been superhigh. Some servicers effectively retain half of borrowers who refinance.”
Meanwhile, “mortgage insurers, which are about 40 times levered, according to our metrics, have massively grown loan books and reduced premiums, even as the agencies’ mortgage underwriting criteria for conforming balance mortgages have been relaxed. Many mortgages are being taken out at a 45% debt to income ratio, and a 95% loan to value ratio. Yet the two-year CDS at 15 basis points is less than on JP Morgan, Citi or Wells Fargo. Delinquencies – which reached 20% in the 2007/2008 crisis – could trigger a blow out in CDS. We do not in fact expect anything as bad as the GFC, as underwriting has not returned to the pre-crisis situation where stated income was allowed, but there is certainly scope for delinquencies and defaults to tick up”.
Given that the portfolio is broadly market neutral, with minimal interest rate and credit spread risk, it might be tempting to apply substantial leverage to these types of trades, but Meneret observes that, “the big lesson of 2008 and even the 2011 crisis is do not be too levered when the markets get choppy. The key change is that dealers do not provide as much liquidity as in the past. We saw a bloodbath in December 2015 as certain mutual funds were forced sellers. Given the record size and leverage in credit markets, and the record size of ETFs, the next financial crisis could centre on credit.” Mill Hill’s leverage, which is calculated in a proprietary way that the manager considers to be conservative, is also adjusted inverse to market volatility. Basis risk is stress tested for worst case loss scenarios while concentration and counterparty exposures are also determined by maximum loss stress testing.
We are emotionless in our sub strategy allocations. If the models indicate better relative value in CLOS, as they do now, we will increase the exposure.
David Meneret, Founder and CIO, Mill Hill
The allocations to the credit sub-asset classes move around within fairly wide bands, according to the opportunity set. “We are emotionless in our sub-strategy allocations. If the models indicate better relative value in CLOs, as they do now, we will increase the exposure,” says Meneret. Thousands of potential trades are monitored using Mill Hill’s quantitative, fundamental and technical three stage process. Proprietary quant models flag up potential valuation anomalies in corporate and structured finance. One key indicator calculates the aforementioned Mill Hill Basis monitor, measuring average basis between typical longs and shorts, and another calculates a credit version of the VIX volatility index, which is at a very low level currently. “But fundamental analysis is needed to confirm that the quantitative valuation gaps are really attractive, and this, in turn must be underscored by technical analysis which limits downside and identifies a specific catalyst for capturing or closing the valuation gap,” stresses Meneret. The process is a hybrid of systematic and discretionary. “It quantifies as much as possible but also applies discretion based on years of experience analysing complex asset classes, such as CLOs, where recovery rates could be zero, and exogenous inputs such as legal settlements need to be heeded,” he points out.
Mill Hill follows a data-driven, forward-looking approach because Meneret has seen that, “a backward-looking approach misses every paradigm shift. In 2008, investors thought sub-prime was great value because they extrapolated from a history of house prices having only ever gone up. After the housing market crash, CDO investors worked on the assumption that default rates would stay high and prepayment rates would stay low forever”.
Meneret has been ferreting out these opportunities for 13 years, since October 2006, previously at Macquarie, and before that, at UBS. Several team members – Senior Portfolio Manager, Gaurav Singhal; Chief Risk Officer and Head of Research, Hongwei Cheng; and COO, Robert Perdock, have worked together for most of this period. Meneret belongs to what is one of France’s biggest exports: the diaspora of financial engineers. His first finance role was in aircraft finance for French bank Credit Agricole, where he built models to value aircraft bonds, which were a complex asset class because the volatility of aircraft values discouraged some lenders from participating in the market. Moving to the UBS principal trading desk he then applied a similar approach to modelling sub-prime collateralized debt obligations, and again to sub-prime RMBS. By 2008, he felt ready for a pure buy side role and moved to Macquarie where he spent seven years co-managing up to $670m in today’s Mill Hill strategy. It was at UBS that Meneret first came into contact with Nick Pepe, Partner, Managing Director, Head of Business Development and Investor Relations who was then working on the sell side at Citi. “The short sub-prime trade has been made infamous by the movie The Big Short but Meneret’s approach to it was distinguished by having shorted the most senior, AA and AAA rated tranches, whereas other managers had shorted more junior tranches,” says Pepe. “Meneret’s focus on the uppermost securities meant that the cost of the CDS insurance was far lower, giving the trade a much more asymmetric and convex payoff profile than if he had paid more to short more junior slices. He funded these shorts with the carry from already impaired tranches of RMBS and ABX Index.”
Meneret soon shifted the focus to trade the 2009 credit recovery, maintaining a market neutral approach. He was not the first mover, as he judged CLOs to be too rich relative to high yield at one stage. But by April 2009, AAA rated CLO tranches in the 60s were too compelling to pass up. Three months later, as the loan market continued to rally, he moved into CLO equity.
The 2011 European sovereign debt crisis – and the liquidation of Maiden Lane CLO – threw up some opportunities for both European and US asset backed securities and associated structured credit, but Meneret naturally used relative value trades rather than outright ones. Seasoned, mezzanine, sub-prime RMBS looked cheap relative to ABX and CMBX, and the two re-converged by 2013.
“We are always looking for convexity on the upside or the downside. The balance between trades geared to upside and downside convexity varies over time, and in late 2019, we think many risk factors are pointing towards higher volatility,” says Meneret.
“We expect that the next crisis will be sparked by liquidity, which is already poor in parts of the credit markets, and this could throw up opportunities for our contrarian approach. Dealer inventories are now down to less than 0.5% of the corporate bond market, which indicates the constrained liquidity situation,” he says. The inventory is now being held in mutual funds with daily liquidity or ETFs which trade throughout the day. As much as one third of the high yield market is now based on passives exposures, and in 2018 passive bond funds saw net inflows of $116.57 billion, nearly six times the $20.89 billion for active funds,” says Meneret, referencing Morningstar data. This means selling could be fairly indiscriminate, throwing out babies with bathwater.
Meneret expects that bespoke credit tranches, or “synthetic CDOs”, could trigger the next credit rout, and there are plenty of catalysts. “Limited dealer liquidity is well understood, but the unwind of bespoke structured credit backed by CDS is perhaps less so. Some bespokes are very levered, up to 80 times to the underlying portfolio, and a recession or increase in defaults could spark panic selling. One deal structured a few months ago looked okay at the time but there are already problems apparent at some of its constituents – Transocean, US Steel, and Mallinckrodt, for instance. As usual, the credit ratings agencies are behind the curve.”
The life insurers that Mill Hill is short of could also unsettle the calm of the credit markets. “When insurers get downgraded, they may become forced sellers, as we saw with Genworth in 2014. But there is no need for the insurers to start selling to create a credit market dislocation. If they merely stop buying, as they temporarily did in December 2018, this could create a rout, as they now account for 25% of the US corporate bond market,” says Meneret, referencing data from CreditSights, which finds that life insurers own 21% and P&C insurers own 4% of the USD 12.6trn corporate bond market.
The September 2019 spike in overnight repo rates to around 10% – the sharpest jump since the great financial crisis of 2008 – is one symptom of impaired liquidity. “It shows that investors are using a lot of leverage and are willing to take on overnight leverage risk to fund portfolios on a day to day basis. The fact that the repo spike took place mid-month is of particular concern,” observes Pepe. “During the crisis, Lehman and other banks used very short-term repos, lending out securities and taking in cash, to reduce the balance sheet leverage they reported at quarter end. The Fed has taken appropriate measures, but bank balance sheets are not as healthy as the market thinks they are now; the new issue market for bank bonds is quite challenging,” he continues. Meneret clarifies, “The repo maturities that Mill Hill uses – between one and three months, with several different counterparties – were not impacted by the spike in repo rates.”
The good news is that positioning for higher volatility need not be expensive. “Credit volatility is also cheaper than at any time since 2007, and could provide significant upside. Though Mill Hill’s portfolio is long volatility, it also has slightly positive carry. We have some very convex short trades, such as owning CDS on BBB rated bonds that are likely to get downgraded in a recession. This negative carry is funded by positive carry from callable bonds on the long side that should also be fundamentally correlated to the shorts. You have to pay up to be long of volatility in equities, but in credit you can combine long volatility with positive carry thanks to second order correlations. We might lag other funds in a rally, but ideal environments for us are when volatility is high, like in 2016 or 2008,” says Pepe. “Indeed, the strategy targets zero correlation to equities and corporate credit.”
The two big themes of basis trades exploiting valuation divergences, and a long volatility bias, mean Mill Hill could be well positioned for two very different economic and financial market scenarios. A “soft landing” and associated normalisation could see a re-convergence of relative value relationships, while a “hard landing” could occasion a quantum leap in volatility.
Mill Hill has an innovative fee structure that reduces management fees as assets grow. Capacity is estimated at $1.35bn, which seems very precise – but Mill Hill also has meticulous methods of estimating strategy capacity. The capacity target (which is a moving target) seems quite low relative to market sizes traded: $2.6trn for corporate financials; $1.3trn for MBS; $600bn for CLOs and $160bn for aircraft and esoteric ABS. But this is not a buy and hold strategy: “We need to trade a lot to execute the strategy and monetise convexity as basis fluctuates. We calibrate capacity to average daily trading volume for the assets we trade, assuming we can execute the strategy without compromise,” says Meneret. In addition to a diverse institutional investor base, Mill Hill garnered a significant inflow in October 2019 from an allocator who has been monitoring the strategy for three years and who feels that the moment is now opportune.