Speaker: Juan Jose Alonso, Professor of Aeronautics and Astronautics, Stanford
More details to come.
More details to come.
More details to come.
Title: "Optimal Trading"
More details to come.
Speaker: Steven Scott, Senior Economic Analyst at GoogleTitle: Scaling Bayesian learning through consensus Monte Carlo
Abstract: A useful definition of “big data” is data that is too big to comfortably process on a single machine, either because of processor, memory, or disk bottlenecks. Graphics processing units can alleviate the processor bottleneck, but memory or disk bottlenecks can only be eliminated by splitting data across multiple machines. Communication between large numbers of machines is expensive (regardless of the amount of data being communicated), so there is a need for algorithms that perform distributed approximate Bayesian analyses with minimal communication. Consensus Monte Carlo operates by running a separate Monte Carlo algorithm on each machine, and then averaging individual Monte Carlo draws across machines. Depending on the model, the resulting draws can be nearly indistinguishable from the draws that would have been obtained by running a single machine algorithm for a very long time. Examples of consensus Monte Carlo are shown for simple models where single-machine solutions are available, for large single-layer hierarchical models, and for Bayesian additive regression trees (BART).
About the Speaker: Steven Scott is a Senior Economic Analyst at Google, where he has worked since 2008. He received his PhD from the Harvard statistics department in 1998. He spent 9 years on the faculty of the Marshall School of Business at the University of Southern California. Between USC and Google he also had a brief tenure at Capital One, where he was a Director of Statistical Analysis.
More details to come.
More details to come.
Title: The Flash Crash: A New Deconstruction
Abstract: On May 6, 2010, in the span of a mere four and half minutes, the Dow Jones Industrial Average lost approximately 1,000 points. In the following fifteen minutes it recovered essentially all of its losses. This “Flash Crash” occurred in the absence of fundamental news that could explain the observed price pattern and is generally viewed as the result of endogenous factors related to the complexity of modern equity market trading. We present the first analysis of the entire order book at millisecond granularity, and not just of executed transactions, in an effort to explore the causes of the Flash Crash. We also examine information flows as reflected in a variety of data feeds provided to market participants during the Flash Crash. While assertions relating to causation of the Flash Crash must be accompanied by significant disclaimers, we suggest that it is highly unlikely that, as alleged by the United States Government, Navinder Sarao’s spoofing orders, even if illegal, could have caused the Flash Crash, or that the crash was a foreseeable consequence of his spoofing activity. Instead, we find that the explanation offered by the joint CFTC-SEC Staff Report, which relies on prevailing market conditions combined with the introduction of a large equity sell order implemented in a particularly dislocating manner, is consistent with the data. We offer a simulation model that formalizes the process by which large sell orders of the sort observed in the CFTC-SEC Staff Report, combined with prevailing market conditions, could generate a Flash Crash in the absence of fundamental information. Our research also documents the emergence of heretofore unobserved anomalies in market data feeds that correlate very closely with the initiation of and recovery from the Flash Crash. Our analysis of these data feed anomalies is ongoing as we attempt to discern whether they were a symptom of the rapid trading that accompanied the Flash Crash or whether they were causal in the sense that they rationally contributed to traders’ decisions to withdraw liquidity and then restore it after the anomalies were resolved.
More details to come.
More details to come.
Title: Macroeconomic-driven Prepayment Risk and the Valuation of Mortgage-Backed Securities
Abstract: We introduce a reduced-form modeling framework for mortgage-backed securities in which we solve for the implied prepayment function from the cross section of market prices. From the implied prepayment function, we find that prepayment rates are driven not only by interest rates, but also by two macroeconomic factors: turnover and rate response. Intuitively, turnover represents prepayments for exogenous reasons like employment-related moves, household income shocks, and foreclosures, while rate response reflects frictions faced by borrowers in refinancing into a lower rate. We find that the implied turnover and rate response measures are in fact significantly related to macroeconomic measures such as consumption growth, the unemployment rate, housing values, credit availability, and market uncertainty. Implied prepayments are substantially higher than actual prepayments, providing direct evidence of significant prepayment risk premia in mortgage-backed security prices. We analyze the properties of the prepayment risk premium and find that it is almost entirely due to compensation for turnover risk. We also find evidence that mortgage-backed security prices were significantly affected by Fannie Mae credit risk and the Federal Reserve’s Quantitative Easing Programs.
This May at the ICME Xpo, get an up-close and inside look at current research and future plans for ICME faculty and students. ICME is engaged with over 50 faculty from 18 departments throughout Stanford. This is a unique opportunity to see how computational mathematics, data science, scientific computing, and related fields are applied across a wide range of domain areas.
ICME Xpo features an afternoon poster session followed by a series of faculty vision talks and promises plenty of opportunities to connect with ICME faculty and students, alumni, and partners from industry and laboratories.
Sample topics to be presented at the poster session include: robot motion optimization, neural networks, single-molecule numerical analysis, optimization of net present value under uncertainty, motif based spectral clustering, and distributed acoustic sensing.
Who should attend:
ICME Faculty, Students, Staff and Stanford colleagues
ICME's Partners in Industry and National Laborotories
ICME, SCCM, and Financial Math Alumni
More details including registration information, location, and event schedule will be announced soon.
More details to come.
ICME proudly announces the second annual Women in Data Science Conference will take place at Stanford University on Friday, February 3, 2017. 
Sign up for our mailing list to receive announcements when tickets are available: https://stanforduniversity.qualtrics.com/jfe/form/SV_37Y5EPI7bMyLInH
View videos of presentations and interviews with conference speakers from the inaugural conference here:
© Stanford University, Stanford, California 94305