Great White North

Yves  Smith:  She’s pretty sure CDS did it.  Maybe she’s just musing a bit, or maybe she’s building case to indict -and-convict here.

ZH: DB Zwirn

ZH: DTCC Update  $15.1 trillion singlename $12.4 index & index tranches total $27.5 trillion  down $700 billion w-o-w $600 of the $700 billion dip is in index.

David Li 1 Johnny Cash 0

FT on copulas. Johnny Cash gave it up for default correlation.

thedeal.com: CDS holders hindering GM debt restructuring

Infectious Greed: Two Contrarian Views on CDS

Kahneman: article about him and his views on irrationality applied to current crisis.

Fed Dallas: Debunking Derivatives Delrium

Geanakoplos: The Leverage Cycle. Unique discussion on the overall effect of having to post collateral in the mortgage and CMO market from an economic modeling context. Take a look at Geanaloplos Yale home page for more papers, some of which are for lay person consumptuion.  Also remarkable that the Tobin Chair of Economics at Yale doesn’t come right out and say CDS did it, but he pretty much figuratively puts CDS at the scene of the crime and is checking if CDS has a gun permit.

Capital Structure

Milken in WSJ: Why Capital Structure Matters

Salmon: Do CDSs cause more bankruptcies?

Bookstaber: Faster isn’t always better. Funny argument against a strawman no one actually believes in. Think of the quest for higher performance like washing. While everyone agrees that compulsive washing, ala Lady Macbeth, is of limited general benefit, the observation hardly merits use of  the electrons needed to blog it i.e. “dude, did you hear, washing is not always better”. Its sort of too Billy Madison to be offered without some obvious ironic distance.  The operative issue is how to identify code that should be kept competitively fast as well as those programs that figuratively haven’t bathed in an extraordinarily long time. I suppose you could also throw in code that is odiferous/stinky from inception or by design. Perhaps APL interpreted code is concise and  gets along well with some of  the rocket scientists but it hasn’t seen a bar of soap since Alan Perlis roamed the lecture halls in New Haven.  Today it is not unheard of to find executives getting used to the nutty fragrance of essentially free-of-charge high performance Java code  running in their production infrastructure. I have come to expect better insights from Bookstaber – I’m looking for  a blog post clarification  of his position – “In the age of Moore’s Law, slower is harmful, almost everywhere”.

Alavian et.al., 2008  Counterparty Valaution Adjustment.

Eric Rosenfeld LTCM: talk at MIT. Head of Fixed Income Trading at Salomon back in the day.

Accounting: Satam Issue confronting PwC.

Salmon’s Bond Dealer talk about Bistro 

Totally Serial

Assume you are approached for an opinion on the computational performance of piece of software that computes the mark to market, risk, and P&L explanatories for a randomly selected 5 year vanilla single name credit default swap. Lets say the software completes computation in 1 second on a contemporary microprocessor in 2009. Assuming the computation output accurate, correct, and complete what benchmarks can you use to get a ballpark sense of how good the performance of 1 second average elapsed time really is.

Lets do a back of the envelope estimate of all the floating point multiplies and adds required for this credit default swap processing on a single microprocessor core. Patterson gives you that a contemporary microprocessor takes 4 clocks for a floating point multiply and 200 clocks to go to DRAM (an L1 and L2 cache miss). Lets assume we are working with a 65 nm Intel Pentium running at 3.6 GHz and that a floating point add on this core takes 2 clocks. So if my actual CDS computation is about half floating point adds and half fp multiplies then in one second I get over a billion 1,000,000,000 operations theoretically to compute required outputs. If I need to use fp divides, they are expensive at 20 clocks per divide and exponentials (vectorized) cost any where from 10 to 30 clocks depending on the accuracy requirements. Just for round numbers lets go with the billion fp ops per second as an average.

At inception a 5 year OTC vanilla default swap has 20 premium payment dates that are actually fixed to be IMM dates . The net present value of these scheduled premium payments is determined by discounting each by the risky discount factor determined by the cooked credit curve. The net present value of the other swap leg, the protection leg, is determined by integrating the product of default arrival probability and the default payout amount over the term of the premium leg. Since the hazard rate is typically piecewise constant (and the recovery rate constant over the future of the swap) the integral can be closely approximated by a weighted summation of expected protection values at each of the premium cashflow dates, the 20 IMM dates for a 5 year swap. Assuming all fp is double precision then each of the 20 cash flow dates requires 8 bytes so the entire NPV sum requires 160 bytes and certainly fits in the L1 cache (probably in the register file as well). So, lets estimate that all the mtm, deltas, gammas, explanatory evaluations for the single vanilla credit default swap requires 100 mtm evaluations. I don’t think its likely that there are even 30 separate mtm evaluations required, so 100 seems conservative. Thus 100 times we need to sum 20 discounted cashflows and 20 expected payoffs on default. Lets assume the code has not been optimized so you need 300 clocks for each valuation, times 100 evaluations, you need 30,000 clocks to do the valuations if the computation is hitting the L1 cache. Lets assume we don’t actually hit the L1 cache all the time and performance on the evals requires an extra 90,000 clocks for memory waits, for a total of 120,000 clocks. That accounts for 120 microseconds (correction: its 40  microseconds but certainly less than 120 microseconds so the remaining estimates should hold) of compute time so the balance of the time must be spent cooking curves.

Lets assume  this code has to cook the underlying credit curve once for each of the 100 evaluations above. Again it seems unlikely that the par curve and various perturbed curves for computing the first and second divided differences and the scenario curves will amount to even 30 cooked curves. We cost it at 100 cooked curves to be conservative.  Lets assume the bootstrap process has to fit five annual par spreads (although typically only 3 of the five are marked).  Furthermore lets assume that the fp operation count of fitting the 5Y spread is the same as the 4Y, 3Y, 2Y, and 1Y spreads (although they obviously require fewer fp ops). Fixing the 5Y point involves some root finding algorithm for a fairly smooth and well behaved function. Presumably you can use Brent or even simple bisection at the cost of say 5 full mtm evaluations  at 300 clocks a piece. These 1500 clocks need to be executed at each of the terms 1Y ( I know not really) , 2Y, 3Y, 4Y, and 5Y for a cost of 7500 clocks. For 100 cooked curves that is 750,000 clocks. Lets assume the L1 cache miss induces a factor of 10 memory wait state penalty so the cost is 7,500,000 clocks.

Now we are talking about using some clock, so the curve cooking + the evaluation  comes to 7,500,000 + 120,000  or 7,620,000 clocks . Out of the 3.6 billion clocks available per second we have managed to use less than 0.3% of them to retire the required operations with tons of wasted clocks to buffer our estimations.  Through this back of the envelope estimation we can find  about 3 milliseconds of work to complete the whole computation end to end on a contemporary microprocessor. So that one second elapsed time performance we started off evaluating isn’t looking so much like A-list performance. Its missing normal performance by roughly a factor of 300. It appears to perform like competitive code running on a microprocessor from the year Titanic won the Oscar for best movie (1997)

BTW to eval 100 OTC trades – full mtm, risk, and P&L explanatories on the  same credit curve looks like about 20 milliseconds using these estimates. The ball park time for the time required to run mtm, risk and explanatories on all the non terminated credit default swaps existing worldwide today: Assume 50,000 different curves and 10,000,000 non terminated CDS then its 50,000* 7.5 ms + 10,000,000* 0.12 ms  about 26.25 minutes. Lets call it an hour because there are some OTC CDS with terms greater than 5Y and we need to leave some more buffer for general programmer apathy/cluelessness.  But it would take a code rewrite;  you’d have to get one of those 65 nm Intel microprocessors; all the trades from DTCC; and all the marks from MarkIT. On the other hand you could run it on your home computer after work, before dinner, and if you cannot afford a new computer but you do all the other stuff on your old PC-clunker; I’m thinking 2.5 hours and you’re done with the world’s overnight credit batch in time to catch the end of  South Park w. the kids, I’m totally serial.

No ILP for You

These look like the slides from Patterson’s 06 Stanford talk on ITunes.

Moore’s Law leads to Parallel Progamming Panic

Patterson IEEE 2008 talk slides - really very clever and probably has the virtue of being true as well.  We are  going to spend a chunck of time tracking down the Patterson, Dally, Hennessy references on parallel programing. I suspect U Washington, Google, Amazon, MIT, and CMU are going to be solid sources as well.

 

Taibbi  at Trueslant.com

iPhone reachable sources for F2B CDS

Active Sources for iPhone reading on the train

EcoFin/Trading: ACT; ZH; Baseline Scenario; 

Inventory: DTCC;  ISDA; Clearing Corp; Information Week; Slashdot; Bobsguide

Data: Bloomberg; Moody’s; Markit; FPML; DHT; DBMS2; The Database Column; FTS; 

Analytics: ISDA Std CDS model; Markit Calculator;  Patterson; Demmel; MapReduce; Hadoop; 

Risk/Regulatory: Gupton; Risk Metrics; OCC; OTS; SEC; FSA;

Accounting: Credit Slips;  Accounting Onion;  Accounting Web;

Merton movie

Top shelf Merton 60+ min. talk on Financial Innovation and the Crisis via Infectious Greed here. We could be turning the corner to higher quality public discussion of the genesis of the crisis , witness Blankfein and Merton. Will the  the clowns respond or make an effort to come back and reiterate their positions? Stay tuned.

The Landscape of Parallel Computing Research: A View from Berkeley 

Lawrence Livermore: Introduction to Parallel Computing ; Training

Texts on Parallel Computing via Demmel here 

slashdot

 

GS Speech

Blankfein’s speech to the council of Institutional Investors spring conference in Washington   here

The 10 Principles are ready and published  in FT  here . Yves Smith says it is a must read and Felix Salmon says it is a listicle of utter genius. I’m telling you NTT is Peter Sellers character with a copy of Copleston next to the keyboard, stochastic calculus class in the past,  and math PhD diploma on the wall. The screenplay almost writes itself.

Sunday Reading

Brookings: The Financial Crisis: An Inside View Swangel 

24/7 Wall St: 25 best Fin Blogs

Fortune: FASB fair value accounting ruling

FT Alphaville: FAS 157-e

Making a list – The idea here is to get a set of qualified blog/online sources that are on target for covering the standard Front to Back topics: Trading, Inventory, Data, Analytics, Risk, and Accounting. The Trading, Investment, EcoFin blog space has long been well covered and highly competitive. The Nobel prize winners are slinging prose/multimedia with the aspiring academics, hedge fund legends, and Gonzo journalists.  The action is so hot and contested that it registered on Michael Moore’s radar. He reportedly has arrived, Predator(1987)-like, to partake in the ultimate hunt. The modest goal here is to factor out several interesting topics including: investment advice, political speculation, legal/regulatory environment, and macroeconomics to get to solid information sources on how the basic functional components of a broker dealer/hedge fund work. If  you allow the analogy of the trading desk/broker/dealer as a “very” elaborate cash register, this is a way to collect the data to describe how the fancy cash register works.    

1. A Credit Trader: The Monoline Delusion
2. Zero Hedge
3. Market Movers 
4. Clusterstock
5. Naked Capitalism
6. FT Alphaville
7. Dealbreaker
8. Alea
9. The Baseline Scenario
10. The Big Picture

Bubble: Calculated Risk; Credit Slips; The Aleph Blog; Seeking Alpha; Rortybomb; Infectious Greed; Hempton; Derivative Dribble; Across the Curve; Accrued Interest; Financial Times; Abnormal Returns;