Felix Salmon, Reuters, The truth about Blackstone and Cordere, here.
But that number is gross revenue, not profit. The profit on Blackstone’s CDS position can be looked at as being the difference between that payout, on the one hand, and the amount that it spent buying the CDS in the first place, on the other. (Although in fact, as we’ll see, it’s more complicated than that.) Unless we have some idea of Blackstone’s cost basis on this trade, we have no idea what its profit was. Bloomberg, however, seems to simply assume that Blackstone’s cost basis for the CDS was zero — that it managed to accumulate all that insurance without paying anything for it whatsoever.
To be sure, Blackstone are smart operators, and I don’t doubt that they’re making a profit on this trade. But we really have no idea how big that profit was.
Matt Levine, Bloomberg, Blackstone Made Money on Credit-Default Swaps With This One Weird Trick, here. CDS valuation has always had these large information asymmetries which make it more expensive to be a market maker on any of the proposed CDS exchanges. I don’t think exchange traded vanilla USD interest rate swaps would have this problem to the same degree.
Really, the only reason to cover this story is its majestic beauty. Which is a great reason to cover it, don’t get me wrong; it’s just that aesthetic appreciation of clever derivatives trades is sort of a specialized niche. Certainly “The Daily Show” didn’t muster much admiration and instead spent seven minutes criticizing everyone else for not covering the story. This is wrong. This trade is so lovely that the proper reaction is to love it and cherish it and hold it close to your heart, not to complain that nobody else does.
There’s one other reason not to worry unduly about this trade, and I hesitate to bring it up, but:It’s not really as bad as it looks. I mean, yes, it is very, very clever. It achieves the second-highest goal of any financial engineering, which is to create genuine value for both parties to a transaction (here, Blackstone and Codere) by taking that value from some third party who’s not in the room.3 So that is great. As Blackstone spins it:
We love Jon Stewart and he continues to be one of the funniest people on TV. But the somewhat boring truth is that we cooperated with Codere and its advisors to save it from bankruptcy or liquidation. We provided capital when no one else would, which allowed the Company to live and fight another day.And they could provide that capital efficiently because they took some value from their CDS writers.
Marco Avellaneda, NYU, Algorithmic and High-frequency trading: an overview, here.
Algorithmic trading: the use of programs and computers to generate and execute (large) orders in markets with electronic access.
Almgren and Chriss, NYU, Dec 2000, Optimal Execution of Portfolio Transactions, here.
We consider the execution of portfolio transactions with the aim of minimizing a combination of volatility risk and transaction costs aris- ing from permanent and temporary market impact. For a simple lin- ear cost model, we explicitly construct the efficient frontier in the space of time-dependent liquidation strategies, which have minimum expected cost for a given level of uncertainty. We may then select op- timal strategies either by minimizing a quadratic utility function, or by minimizing Value at Risk. The latter choice leads to the concept of Liquidity-adjusted VAR, or L-VaR, that explicitly considers the best tradeoff between volatility risk and liquidation costs.
xcelerit, Benchmarks: Ivy Bridge vs, Sandy Bridge for Financial Analytics, here. Folks are reporting that the performance jump from Sandy Bridge to Ivy Bridge on this MC code is mostly explained by the increased core count 12/8 =1.5. It is a little uncomfortable not knowing how the code is compiled but the relative figures make sense also in light of the fact that Sandy Bridge and Ivy Bridge have the same AVX architecture 8 DP FLOPS/cycle. This will change dramatically with Haswell AVX2 and FMA which should double the flops per cycle on a suitable MC code while keeping the core count flat to Ivy Bridge.
The table below shows the speedups for different numbers of paths, comparing the Ivy-Bridge processor vs. the Sandy-Bridge processor:
Paths Speedup Ivy-Bridge vs. Sandy-Bridge 64K 1.15x 128K 1.25x 256K 1.34x 512K 1.4x 1,024K 1.48x
As can be seen, the Ivy-Bridge processor gains significantly compared to the Sandy-Bridge, reaching 1.5x speedup for high numbers of paths. This is in line with the increase in the number of cores from 8 to 12 per chip. The benefits of the new Ivy-Bridge for financial Monte-Carlo applications can be clearly seen here.
Michael J. Moore & Dakin Campbell, Bloomberg, Wall Street Sweats Out Volcker Rule Impact on Revenue, here.
Wall Street banks, which already shut proprietary trading units that helped fuel record profits, are girding to learn next week how much revenue the Volcker rule may cut from the $44 billion they say comes from market-making.
With U.S. regulators scheduled to vote Dec. 10, the largest firms are getting little detail about the final terms of the Volcker rule’s ban on proprietary trades, and still have basic questions about what kind of market-making will be allowed, said three senior U.S. bankers. They’re also wondering whether they’ll have to change practices or curtail business in some less-liquid markets, the bankers said.
Matt Levine, Bloomberg, EU Is Shocked That Banks Colluded on Libor, here.
So the banks got together and decided: Let’s create a composite of our borrowing costs and all sell swaps against that composite. We’re all talking to each other anyway as we go about borrowing from each other, so let’s all just write down how much we’re paying to borrow, send our costs to a trade association, take a trimmed average, call it the London interbank offered rate, and write all our swaps against Libor. We can even set up a different trade association to make sure that we all have the same documents for our swaps, so that all our swaps work the same and use the same rate that we all more or less agree on.
So they did that, and it was great. I mean, it was, for them. You can complain because Libor has fallen into some disrepute of late, and they can complain because “fallen into disrepute” really means “has racked up some enormous fines for Libor banks,” but I don’t want to hear it from any of you. U.S. banks — U.S. banks alone – made $2.8 billion just last quarter from trading interest rate derivatives. That decision to have a standardized thing that they all agreed on as the basis for those derivatives worked out just plain great for them.
Albanese, et.al., Jan 2011, Coherent Global Market Simulations For Counterparty Credit Risk, here. CVA floating point hacks.
To explain how greater performance can be achieved by doing more calculations, let us recall a few traits of the silicon economics affecting microchip and board designs. In recent times, there was in fact a radical shift in this landscape.
It used to be that:
(i) Computing capabilities were limited by the ability of ALUs to execute floating point and integer arithmetics
(ii) Memory was expensive and a scarce resource
(iii) Most algorithms were single-threaded and parallelism was best brokered transparently by middleware layers dispatching jobs to large grid farms
(iv) Code was best written in native C++ optimized in such a way to speed up the execution of a great variety of bespoke algorithms.
Although these practices are still widespread in the transition period we are living, the underling technology has now shifted quite radically.
(a) Nowadays, it is relatively cheap to populate microchips with highly capable ALU cores. The 8-socket CPU boards of the emerging generation entail as many as 80 cores capable of hyperthreading in the case of Intel or 96 cores in the case of AMD. Even more extreme ALU counts are seen in the GPU space where the AMD Firepro GPUs have 1600 cores and nVidia Fermis have 512.
(b) Memory is relatively cheap and readily available up to terabyte scale, thus enabling single node technology for portfolio processing as a viable alternative to grid computing.
(c) The clock frequency and bandwidths of data paths are not keeping pace with the compute power of ALUs and the massive memory available, rendering the memory bottleneck tighter than ever within the bounds of cost effective designs.
(d) Vastly different microchip architectures have emerged, including SIMD multiprocessors with up to 16-32 data registers located in discrete GPU parts as in the nVidia Fermi and ATI Firestream, the multicore MIMD designs on CPU boards by Intel and AMD and the emerging MIMD-SIMD hybrid fusion architectures, the Intel Sandybridge and AMD Booldozer.
(e) MIMD and SIMD designs are characterized by radically different threading models: SSE2/SSE3/AVX primitives rule with CPUs while the lightweight, no-frills threading models in CUDA/OpenCL are used for GPUs.
(f) Cache hierarchies for MIMD architectures are complex and involve up to 2 MB per core. GPUs instead are nearly cacheless except for a modest amount of shared memory located on individual SIMD microprocessors.
(f) On the programming language side we see the merit a bifurcation away from catch- all C++ coding. On the one hand, the variety of architectures motivates a revival of interest in low-level optimization of basic building block algorithms. On the other hand, the complexity of multi-threaded orchestration in shared memory designs using large scale in-memory processing motivates the use of higher level languages. Features such as garbage collection, managed thread pools and support for service oriented architectures are in fact essential for complexity management.
Nassim Taleb is a former trader who wrote a textbook on option and market making, and then became more philosophical in his best seller Fooled by Randomness, and now in The Black Swan. His big idea is that sometimes, unexpecting things happen: countries dissolve into anarchy, wars start, unknown authors become famous. His secondary ideas are variations on this theme, that people, especially experts, are generally biased, overconfident, and rationalize past event so they appear deterministic. Stated baldly, these assertion are hardly novel but true enough, and one can argue about their relevance in various cases. As a highly popular presentation of ideas near to my interests and vocation, I think it is worth critically examining if there is anything to his particular contribution to the literature on cognitive biases or social failures. My conclusion, in short, is no.