Quant credit trading

Kitno · Post by **Kitno** » Wed Nov 25, 2015 1:17 pm

There's a hugely insightful comment in the FT in "Bond market blues sneak up on asset managers" from State Street how bond-based ETFs are the solution to manger liquidity problems.

Failing to address the underlying security problems, failing to understand their risk characteristics and their lack of uniformity and packaging into something it is not has worked so well before.

EDIT: I'm going to qualify my above comment as applying to multi-issuer, corporate debt. A similar structure (non-ETF) structure works/worked well during the Greek crisis with GGBs. You could use the same structure for an issuer's credit (with some a small number of caveats) to improve liquidity & depth in some circumstances.

gentinex · Post by **gentinex** » Sat Dec 19, 2015 9:11 pm

I have a fair amount of experience in this space. As others have mentioned, the main issue has been liquidity.

The most natural product to trade quantitatively would have been single-name CDS, as this is clean (pure credit play, no rates, vol or other non-standard corp bond features), but it seems that the liquidity here was propped up by synthetic CDO hedging, and now that synthetic CDOs are basically dead, single-name CDS have been fading as well.

So more recently, people have been moving to the cash space. While there are a non-trivial number of corp bond with unusual covenants and such, there is enough issuance in standard stuff that this could be systematically traded, if again there was liquidity. Here, the main problem seems to have been dealers' reduced balance sheet capacity in the post-crisis regulatory environment. The buy side is very interested in an alternative non-dealer-focused market structure, but no one seems to have come up with a satisfactory solution yet. (Though not for lack of trying - cf. the failures of GSessions and Aladdin (though Aladdin seems to now be getting revived under TradeWeb) and some current startups (see e.g. TruMid and Electronifie.) My guess is that MarketAxess or TradeWeb may end up being the dominant platforms for limit-order-book-like credit trading, but it may take a while to get there (and has already taken longer than many people expected).

The one growth area in terms of liquidity has been the indices - both on the CDS (CDX IG / HY, Itraxx Main / Crossover) and the cash (LQD, HYG, JNK, iBoxx TRSs) sides, and also in credit index options (though the main liquidity here seems to be around regulatory arb rather than quant trading, in that banks have discovered that buying rolling short-dated credit index options is a cheap way to reduce balance sheet for reg cap calcs, and the buy side is taking the other side of this).

Citadel used to be more involved in the single-name space (bond-CDS basis trading, credit-equity trading using Merton models and such, stat-arb-like approaches which involve constructing credit portfolios from factor models), but has been scaling back over the last year or two. My impression is that they're large enough and have good systems in place to deal with the liquidity issues - i.e., their models know which names are and aren't liquid, and can get reasonable execution on a large portion of what they want to trade systematically - but that the returns were still unattractive compared to their core business in converts. I think they are still active in credit indices, though I think their play here is to be an alternative market maker.

There are a handful of quant funds involved in indices, though more as part of a cross-asset momentum strategy than as a purely credit-focused strategy.

Bond-CDS basis trading has always been around, but again, with single-name CDS liquidity drying up, the future of this trade is uncertain.

In recent years there have also been a number of small shops involved in bond ETF arb (trading ETF against single-name bonds), but this has also become a less interesting space due to (1) dealers are now smarter about moving single-name bonds in sync with indices, so there's less of an arb; (2) the create-redeem mechanisms now require delivery / receipt of more bonds than previously, so harder to execute with decreasing bond liquidity.

Kitno · Post by **Kitno** » Wed Dec 23, 2015 10:37 pm

Hi getinex, a very insightful post.

You touch on the various trading platforms too, which warrants a thread of its own. I will inaugurate one under "Trading". While we're on it I'll start a separate one about journos I've come to respect.

I do have to say though that limit-order credit trading promises a poor future.

The easiest non-balance sheet/trading metholody change that would best support liquidity is documentation standardization and an industry strict de-limitation by market of 'benchmark' and reverse inquiry/MTN. Elimination of ambiguity is helpful. Not quite appropriate nowadays but a few years ago 'should the flow desk bit on a $300MM Boeing issue?'. No it shouldn't. Certainty breeds confidence.

baghead · Post by **baghead** » Thu Jan 28, 2016 9:38 pm

Bloomberg - When Systemic Trading Comes to Credit Markets

Could be a lame attempt to revive an otherwise dead product. But I'm out of the market for a few years now...

polysena · Post by **polysena** » Thu Jan 28, 2016 9:58 pm

I am surprised.. so many correlation desks closed down, how can this be booming?

Kitno · Post by **Kitno** » Fri Jan 29, 2016 12:42 am

An interesting article. I don't contest the first few paragraphs.

On a side note linked to the article: I've found it almost amusing how flavor of the month is equity indices following front-month oil. I remember in the prior crisis when it was the bund future. Oh well. Live and learn.

rowdyroddypiper · Post by **rowdyroddypiper** » Sun Mar 13, 2016 10:52 am

@polysena: I think you've answered your question a bit already. Old habits die hard and in this case old soldiers don't fade away. If your background was in credit, but not like a bottom up fundamental underwriting capacity there's a set of skills that can be applied to these credit vol kind of strategies. Trade your Gaussian copula for Black-Scholes and hope to print some trades while the market exists.

Edit: addressing Polysena

Edit^2: as with all things unserious yet terrible I suspect the Phrench are behind it.

Cheng · Post by **Cheng** » Mon Mar 14, 2016 9:01 am

as with all things unserious yet terrible I suspect the Phrench are behind it

Big Smile

Good to see you here, RRP.

rowdyroddypiper · Post by **rowdyroddypiper** » Tue Mar 15, 2016 10:08 am

Thank you! Good to be back. No idea why I took a hiatus in the first place.

polysena · Post by **polysena** » Tue Mar 15, 2016 11:18 am

@rowdy thank you; you are certainly right, last bravado?... and of course some jurisdictions did not implement the same rules at the same time..

Annex question: I always wonder is a "market" consisting in one mm desk and perhaps many HF buyers/sellers still be sustainable (monopole/market?).. there might be arguments to having just one mm is not a problem, but for indices and price discovery, isn't one mm just too small?

Any thoughts anybody? Poly