Pink Iguana

Home » Trading

Category Archives: Trading

Capital Structure Arbitrage

David, Deus Ex Macchiato, If it doesn’t work as a hedge fund strategy, try making policy with it, here.

Capital structure ‘arbitrage’ is largely discredited as a hedge fund strategy for the rather good reason that a lot of people lost a lot of money on it. An arbitrage, remember, is supposed to involve a risk-free profit. But using the Merton model or its variants to ‘arbitrage’ between different parts of the same companies’ capital structure didn’t work very well — or, at least, it worked well until it didn’t. One of the problems (aside from various liquidity premiums embedded in prices) is that the first generation of these models assumed that the value of a firm’s assets follow a random walk with fixed volatility: which they don’t. In fact it has been known for over two decades that the PDs backed out from Merton models are far too high, something that KMV try to fix with some success at the cost of what might kindly be termed ‘pragmatic adjustments’. Now there may well be capital structure arbitrage models which don’t have these first generation problems and don’t involve arbitrary adjustments, but they are not well known (not least because if you had one that worked, you would want to use it to trade rather than to burnish your academic credibility).


High Frequency Trading Technology Stack

Vlad Tenev, Forbes, Quora, What Is The Technology Stack Like Behind A High-Frequency Platform, here.  Odd that this comes out in Forbes, right? 10 microsecond one way gateway latencies on vanilla x86 gear with kernel bypass is standard w. Linux. This year (2014) you will see (as standard) one way sub microsecond native protocol gateways hooked to 10GB Arista ports. Getting the cash equity latencies minimized between Colos is still a challenge: looking for Weehawken to Mahwah at ~140 mics, Weehawken to Carteret @ ~100 mics, Carteret to Mahwah @ ~210 mics, and Aurora to Carteret @ sub 5 millis. Strike Technologies is reporting 4.3 millis Aurora to Carteret. Mr. Tenev is spot on about the servers being on par with nice inexpensive gaming rigs – hence the DynaRack opportunity. Maybe that will heat up after folks get their Colo to Colo latencies nailed down/deterministic. I would imagine that once you know how you want your optimized ALGO to execute, then you are going to have a pretty good idea about how you want to configure DynaRack to run the optimized ALGO code. Would not be shocking if there were a  patent or two you could push through on the DynaRack buildout and configuration.

How about a service where prior to the market open you are offered the chance to test any of 1000 randomly selected sessions to say the BATS matching engines? If there is any variation in the session latency on any of the sessions that you can detect, then you can bid for exclusive use of that session for the remainder of the trading day. Nominally all the exchange sessions should have similar latency performance but the exchange may be rolling out upgrades or new hardware, let’s just check to be sure. I’ll bet there is some reproducible latency variation. Need a name for this new service -call it DynaSession:

Hey, Joe how’d you do in the DynaSession auction this morning?

Oh, not so bad, got six sessions with -2.3 mics to the mean apiece for 100 bucks.


For the lowest possible latency, you would plug the physical cross-connect directly into a network card on your trading server. However, this is seldom practical, as you will likely want multiple machines for failover/warehousing and will often have dozens of exchange cross-connects. So, we recommended a low-latency switch, such as the 24-port Arista. This added 350 nanoseconds (port-to-port) latency. There was also a Blade Networks (now IBM IBM +1.99%) switch that used the same Fulcrum ASIC and was lower priced.

Next , your server would need a low latency network card with a kernel bypass driver. We recommended Myricom (10G-PCIE2-8C2-2S) for UDP traffic and Solarflare (Solarflare Flareon Ultra SFN7122F Dual-Port 10GbE PCIe 3.0 Server I/O Adapter – Part ID: SFN7122F) for TCP.

Both of these cards provide kernel bypass drivers that allow you to send/receive data via TCP and UDP in userspace. Context switching is an expensive (high-latency) operation that is to be avoided, so you will want all critical processing to happen in user space (or kernel-space if so inclined).

We compiled and provided our clients with a lightweight custom build of Linux (Gentoo), which we maintained in-house.

We pinned our trading software to dedicated cores and ensured other processes didn’t run on those cores (again to avoid context switches). We also disabled local timer and other interrupts on those cores.

The platform logic code was all event-based and written in C. We developed our own lockfree data structures (FIFOs and Ring Buffers), and each platform thread basically ran as an infinite loop polling its own input FIFO.

The end result was a system with less than 15 microseconds of latency (wire to wire). In terms of cost, the low latency switch ran about $15k, and you could put the server together for less than $5k, on par with a nice gaming rig.

Radio Frequency

Andrew Tangel, LA Times, Firms race to transmit Wall Street data at nearly light speed, here. Who told about the DynaPie dirigibles?

It could be the final frontier in the financial industry’s tech arms race as the fastest traders scramble to get even faster, and as regulators mull over ways to prevent technology breakdowns on Wall Street. Industry insiders say that some firms are toying with lasers and high-altitude balloons.


Knight’s Landing at SC13 and Leveraged SCDOs

Andy Patrizio, IT World, Intel changes the whole supercomputing game with Knight’s Landing, here.

Much more important is what else it takes away. Knight’s Landing will erasing the memory buffer and PCI Express bus that sat between the CPU and main memory and the coprocessor chip and frame buffer memory in the Xeon Phi card. Now that applications run entirely natively instead of offloading the data sets to the coprocessor, all of that latency goes away.

Now you will have both scalar processor cores and vector processor cores on the same chip sharing access to unified memory. This is huge. A fair amount of time has to go into offloading data sets from main memory to the frame buffer memory of the co-processor and then back to the CPU and main memory. It’s why Nvidia had to come out with the CUDA language, because plain old C++ or Java wouldn’t work.

Matt Levine, Bloomberg, Welcome Back, Leveraged Super Senior Synthetic CDOs, here.

You can tell that leveraged super senior synthetic collateralized debt obligation tranches are fun because they are called leveraged super senior synthetic collateralized debt obligation tranches, and anything with that many words in its name is up to something. And in fact LSS CDOs were popular prior to the financial crisis, got various people in various kinds of trouble, and more or less vanished.

But now Euromoney is reporting that Citigroup is trying to market them again, with a slight modification that might get people into less or at least different kinds of trouble, though it is far from clear that anyone will be interested.

Bond Liquidity and Classy Dancing

Felix Salmon, Reuters, When bonds don’t trade, here. Lots of fixed coupons folks cannot easily trade out of. Maybe they will use swaps?

Loss of Shock Absorbing Capacity.jpg

This chart doesn’t just cover Citigroup, it covers all bond broker-dealers. They massively increased their inventory of bonds during the 2000s bubble — but so did everybody else: total credit assets were raising substantially over that period. Then, after the financial crisis, came the great divergence. Broker-dealers retreated from the market, even as investors continued to seek the safety of bonds. So while broker-dealers were about half the size of the credit mutual fund industry in 2007, according to the quantity of assets they owned, today they’re only about 1/20th of the size. And those broker-dealers are still the only real liquidity providers in the market. If you want to buy or sell a bond on the secondary market, there’s really only one way to do it: phone a bunch of broker-dealers, ask them to make you a market, and either accept the best price you find, or don’t.

Lee’s article makes a very strong case that the only way out of this problem is for buy-side institutions to start trading directly with each other, since the broker-dealers have enough to be able to provide good service only to their very best clients. But neither of the two buy-side bond market giants (Blackrock and Pimco) seem to have been able to make such a system work, and although the MarketAxess system is growing fast, there isn’t going to be any fundamental change unless and until bond investors start making buy/sell markets of their own. Which is simply not going to happen: bond investors don’t tend to think in terms of opportunistic trading, precisely because their portfolios are so illiquid. What’s more, the ability to make a two-way market is contingent on the ability to buy one name when you sell another, which is not something anybody can reliably count on being able to do any more.

In other words, we’re living through a vicious cycle: the less liquid the market gets, the less ability there is for anybody to make markets, which in turn just worsens the liquidity problem. And things are only going to get worse still if and when QE goes away.

Dan Amira, New York Magazine, James Dolan Wants the Knicks City Dancers to Be Classier, here. That’s why Lead Farmers have no Knicks on the roster – not enough Classy Dancing. Update: well if Tyson Chandler is out for 6 weeks and there is going to be classy dancing at MSG then the Lead Farmers will take the new starting center for your New York Knickerbockers … Bargs off the wire.

The source said there is a movement to upgrade the routines by the popular Knicks City Dancers to make them less like “regular cheerleaders.”

“We’re in the process of a rebranding of the Knicks City Dancers,” the source said. “We’re looking at making them bigger and better so they’re not just regular cheerleaders.” The Garden source added the plan is to make the routines more “classy,” citing the tap-dance number that was performed Sunday night.

Algos and Ditch Your Index Funds

Dan McCrum, FT Alphaville, Puny human analysts to be crushed by algorithmic steamroller, here.

The world’s most successful hedge fund strategy is both the most secretive and the one that you could never invest in: Renaissance Technology’s Medallion fund. The fund ignores the mainstream finance literature, preferring to scoop up experienced cryptographers and mathematicians required to sign intimidating non-compete clauses, and it is willing to trade signals in the noise that work even if it doesn’t understand why.

According to II, however, new systems on offer to analyse the flood of news headlines and tweets can cost from $5,000 to $20,000 a month, less than the annual pay of a junior hedge fund analyst.

We’re reminded of AQR, the hedge fund firm built upon a former Goldman Sachs prop desk that has become an all-round quant-based asset manager. Each year it offers a $100,000 prize to get an early look at unpublished papers from the world of academic finance.

S.L. Mintz, Institutional Investor, Analysts Beware! A Machine Has Its Eye on Your Job, here.

AQR Insight Award 2014, here. Previous winners, here.

You are invited to submit a paper for the 2014 AQR Insight Award. A $100,000 prize will be awarded to the most important unpublished finance paper that provides significant, novel, practical solutions for institutional investors and financial advisors. The judges will have discretion to select up to three papers to share the prize.

Dan McCrum, FT Alphaville The rate exit, Credit Suisse edition, here.

The Swiss bank has followed the lead of UBS in deciding that core fixed income trading is just too expensive, now that the whole flight to safety trade is over and lucrative over the counter business is dwindling.As the FT reported last month:

Regulators around the world have raised capital requirements in the wake of the financial crisis and this, combined with subdued markets and a move towards central clearing, has prompted banks to rethink how they run their debt trading.

Matt Levine, Bloomberg, Levine on Wall Street: Ditch Your Index Funds, here.

Stock-picking is back

Actively managed mutual funds have been so maligned for so long that I guess they were due for a good year, and so Reuters reports that “Some 57 percent of U.S. funds run by active managers are beating their benchmark indexes this year.” This seems to be attributable to the fact that stocks are moving less together and doing more of their own thing: Implied correlations “have fallen to their lowest since October 2007 after peaking in 2011. “It’s a stock-picker’s market,” as they say, and boy do they ever say that. (Not Reuters, I mean, just the generic “they.” Though Reuters too.) One lesson here is that investing success is built on a series of meta-skills and meta-judgments: Sometimes it is nice to be able to pick stocks (or to pick fund managers who pick stocks), other times wisdom lies in indexing, and knowing in advance which times will be which has obvious value.

Drone-Mounted Lasers

Derek Mead, Motherboard, Drone-Mounted Laser Weapons Are on the Way, here. Need this to protect the DynaPie Trading Dirigibles

The project, called Endurance, is referred to in DARPA’s 2014 budget request as being tasked with the development of “technology for pod-mounted lasers to protect a variety of airborne platforms from emerging and legacy EO/IR guided surface-to-air missiles.” The budget explains that it will be the first application of DARPA’s much-discussed Excalibur laser defense system, which developed lasers powerful enough to use as weapons.

Making Markets in CDS is Hard

Ruhle, Childs, & Miecamp, Bloomberg, Blackstone Unit Wins in No-Lose Codere Trade: Corporate Finance, here.

The unit of Blackstone Group LP (BX) structured the loan in a way that would lead to a payout on swaps it held, according to three people with knowledge of the situation who asked not to be identified because the discussions were private. The contracts were triggered on Sept. 18 after Codere delayed an interest payment by two days to comply with the loan terms. GSO held 25 million to 30 million euros of the swaps, meaning it may have made at least 11.4 million euros ($15.6 million), according to one of the people and data compiled by Bloomberg.


Knight Capital Americas LLC

Alexandra Stevenson, NYT, Knight Capital to pay $12 Million Fine on Trading Violations, here.

On the morning of Aug. 1 2012, Knight Capital, Wall Street’s biggest trading firm, sent out a wave of accidental stock orders – more than four million – that reverberated through the market and eventually resulted in a $460 million loss for the firm.

SEC, In the Matter of Knight Capital Americas LLC Respondent, here. The dead “Power Peg” code activation on a failed production roll distribution, described here, seems real hard to test in UAT. Bet that it’s a good time to talk to Appel about Verified Software Toolchain if you are the GETCO folks who just purchased Knight and are paying the 12 million USD fine. Not obvious to me that the failed production roll distribution process is going to be in scope for VST, but who knows, maybe they thought about that. I suppose you want the gateway, FPGA or otherwise, connecting you to the exchange verified, right? Locking that up could be lucrative and the sales pitch is sort of straight forward, once you read the SEC/Knight Cap. post mortem.

On August 1, 2012, Knight Capital Americas LLC (“Knight”) experienced a significant error in the operation of its automated routing system for equity orders, known as SMARS. While processing 212 small retail orders that Knight had received from its customers
SMARS routed millions of orders into the market over a 45-minute period, and obtained over 4 million executions in 154 stocks for more than 397 million shares. By the time that Knight stopped sending the orders, Knight had assumed a net long position in 80 stocks of approximately $3.5 billion and a net short position in 74 stocks of approximately $3.15 billion. Ultimately, Knight lost over $460 million from these unwanted positions. The subject of these proceedings is Knight’s violation of a Commission rule that requires brokers or dealers to have controls and procedures in place reasonably designed to limit the risks associated with their access to the markets, including the risks associated with automated systems and the possibility of these types of errors.

David Swenson

Dan McCrum, FT Alphaville, You’ll never be a Yale superman, here. I suspect that there is Yale and Ptown cash around for trading IR swaps on SEFs.

Consider the recent decision by the $21bn Yale endowment to cut its allocation to private equity for the first time in eight years, to 31 per cent from 35 per cent.

That shift may simply reflect a lack of opportunities at present, but note the long-term return target which Yale expects its private equity investments to deliver.