All seminars are held in 639 Evans Hall at UC Berkeley, unless otherwise notified.
Upcoming seminarThere are no upcoming seminars.
The seminar will be held at 1011 Evans Hall, UC Berkeley.
Triggered by the introduction of ever stricter accounting and prudential pension fund regulations, a massive shift from defined-benefit to defined-contribution pension schemes is taking place across the world. As a result of this massive shift of retirement risks on individuals, the investment management industry is facing an increasing responsibility in terms of the need to provide households with suitable retirement solutions. Existing retirement products such as target date funds, annuities and variable annuities suffer from a number of shortcomings which make them ill-suited for investors saving for retirement in the accumulation phase of their life-cycle. In this paper, we describe how dynamic asset pricing theory and financial engineering can be used to design scalable mass-customised forms of retirement solutions that can address the specific retirement needs and constraints of a large number of individuals in a parsimonious manner.
Lionel Martellini is Professor of Finance at EDHEC Business School, Director of EDHEC-Risk Institute and Senior Scientific Advisor for ERI Scientific Beta. He is a former member of the faculty at the Marshall School of Business, University of Southern California, and has been a visiting fellow at the Operations Research and Financial Engineering department at Princeton University.
Lionel holds Master’s Degrees in Business Administration (ESCP Europe), Economics and Statistics (ENSAE) and Mathematics (Paris 6 University), as well as a PhD in Finance from the Haas School of Business, University of California at Berkeley. He also recently completed a PhD in Relativistic Astrophysics (University Côte d'Azur) and has been involved in the LIGO/Virgo international collaboration for the observation of gravitational waves.
Lionel is a member of the editorial board of The Journal of Portfolio Management, The Journal of Alternative Investments, and The Journal of Retirement. He conducts active research in a broad range of topics investment solutions for individual and institutional investors, equity and fixed-income portfolio construction, risk management and derivatives valuation. His work has been published in leading academic and practitioner journals and has been featured in major European and global dailies such as The Financial Times and The Wall Street Journal. He has co-authored reference textbooks on topics related to Alternative Investment Strategies, Fixed-Income Securities and Investment Solutions.
Lionel has served as a consultant for institutional investors, investments banks and asset management firms on a number of questions related to risk and asset allocation decisions, and is a regular speaker in seminars and conferences on these subjects.
I examine whether whether low capital levels incentivize banks to systematically originate and hold riskier loans. I construct a novel data set consisting of 1.8 million small business and home mortgage loans, matched to the specific banks that originated them and the capital levels of those banks at the time of origination, and verified to be held on bank portfolios, rather than sold. A one point increase in capital ratios (e.g. from 12% to 13%) is associated with a 4% decrease in the default risk of mortgage loans held on portfolio (from a net foreclosure rate of 2.5% to 2.4%). Bank capital has macro impacts. When considering the average capital of banks in counties during the pre-crisis period, a one point increase in capital levels is associated with a 2.9% reduction in foreclosures during the financial crisis. A five point increase in capital ratios, which was achieved post-crisis, could have prevented at least 430,000 foreclosures had it occurred earlier. These results are robust to bank and time fixed effects and an instrumental variables strategy for predicting bank capital.Read the paper this talk is based on: Ohlrogge_Bank_Capital_8_28_17 Ohlrogge_Bank_Capital_Online_Appendix_8_28_17 Download the slides from this presentation: Bank_Capital_Aug_2017_Presentation
Sparse Dictionary Learning (SDL) can be used to extract narrow factors driving stock returns from a stock returns matrix, provided the returns are generated by sparse factors alone. We describe progress on a variant called Sparse Low Rank Dictionary Learning (SLRDL), designed to simultaneously extract broad and narrow factors for the returns matrix, when the returns are generated by both types of factors.
We formulate the portfolio construction problem as a mean/variance problem which includes a linear term representing an investor’s preference for expected “social return”, in addition to her expected “financial return” of the classical theory. By making various assumptions, we are able to exploit the heterogeneous expectations version of the CAPM to derive an equilibrium model which is an extension of the standard Capital Asset Pricing Model. Among other things, the model implies that, in equilibrium, assets with higher expected social return that is valued by investors will have, ceteris paribus, lower financial expected return. We also present guidelines for practical implementation of this approach to portfolio management.Download the slides from this presentation: Berkeley Presentation 20170912 v1.0
Since the work of Page in the 1950s, the problem of detecting an abrupt change in the distribution of stochastic processes has received a great deal of attention. There are two main formulations of such problems: A Bayesian approach where the change-point is assumed to be random, and a min-max approach under which the change-point is assumed to be fixed but unknown. In both cases, a deep connection has been established to variations of the widely used CUSUM procedure, but results for processes with jumps are still scarce, while the practical importance of such processes has escalated. In this talk we consider change-point detection problems for processes with independent and stationary increments, i.e. Levy processes, as well as some important extensions, such as to processes with a dependence structure, and to the case of distributional uncertainty.
We begin with a survey of machine learning techniques and applications outside of finance. Then we discuss our use of Machine Learning techniques at Rosenberg. Finally, we explore some alternative data sources.
Factor models are used to predict the future returns of a portfolio with known positions in many assets. These simulations yield a distribution of future returns and various measures of the risk of the portfolio. Clients would often like to identify sources of risk in their portfolios, but this is difficult when factors influence the portfolio in nonlinear ways, such as when returns are measured on a log scale and when the portfolio contains nonlinear instruments. We develop a two-step method to partition risk among factors in a portfolio which accounts for these nonlinearities: first, model the relationship between factors and portfolio returns, and second, estimate the risk contribution of each factor as the increase in portfolio risk due to increasing the factor's weight. Both of these steps can be done using nonparametric regressions, which make no assumptions about the distribution of factors or their functional relationship with the portfolio returns. This research was done at State Street GX Labs.
In the 2013-2014 season, the National Basketball League, in conjunction with STATS LLC, implemented a league-wide program to collect player-tracking data for all NBA games. The data feed now provides 25-FPS records of all players' XY coordinates on the court, as well as XYZ coordinates for the ball. This data source has opened up new lines in inquiry into the quantitative analysis of basketball that have previously been hamstrung by a reliance on spatially naive box-score and play-by-play statistics. In this talk I will discuss several projects undertaken by myself and the XY Research group that use newly-available spatial data to work toward answering fundamental questions about basketball. Topics covered will include expected (EPV, or a stock-ticker for a possession), defensive shot charts, the impact of ball movement, and play detection.
Backtest overfitting means the usage of backtests (historical market data) to construct an investment strategy, fund or portfolio, when the number of variations explored exceeds limits of statistical reliability. We show that backtest overfitting is inevitable when computer programs are employed to explore millions or even billions of parameter variations (as is typical) to select an optimal variant. We illustrate this by showing that while it is a simple matter to design a stock fund, based only on a weighted collection of S&P500 stocks, that achieves any desired performance profile, these funds typically perform erratically at best when confronted with new, out-of-sample data. Similarly, we present results of a recent study of market forecasters, most of whom employ some sort of historical market data analysis, and show that few, if any, have a positive long-term record.
Read more about his research: http://www.davidhbailey.com/Download the slides from this presentation: dhb-risk-2017
We analyze the optimal allocation of trades to portfolios when the cost associated with an allocation is proportional to each portfolio's risk. Our investigation is motivated by changes in the over-the-counter derivatives markets, under which some contracts may be traded bilaterally or through central counterparties, splitting a set of trades into two or more portfolios. A derivatives dealer faces risk-based collateral and capital costs for each portfolio, and it seeks to minimize total margin requirements through its allocation of trades to portfolios. When margin requirements are submodular, the problem becomes a submodular intersection problem. Its dual provides per-trade margin attributions, and assigning trades to portfolios based on the lowest attributed costs yields an optimal allocation. As part of this investigation, we derive conditions under which standard deviation and other risk measures are submodular functions of sets of trades. We compare systemwide optimality with individually optimal allocations in a market with multiple dealers.Download the slides from this presentation: submodular risk allocation_ghamami_10_17
It has been recently shown that rough volatility models, where the volatility is driven by a fractional Brownian motion with small Hurst parameter, provide very relevant dynamics in order to reproduce the behavior of both historical and implied volatilities. However, due to the non-Markovian nature of the fractional Brownian motion, they raise new issues when it comes to the risk management of derivatives. Using an original link between nearly unstable Hawkes processes and rough volatility models, we explain in this talk how to price and hedge options in the rough version of the Heston model. This is joint work with Omar El Euch.Download the slides from this presentation: Pres_Rosenbaum_Berkeley_071117
The Market is a consensual hallucination that commands attention by wielding its Invisible Hand. In this talk we will examine the ways that Adam Smith’s 250-year-old appendage makes itself felt – positioning, trading, and hurting herding – and their implications for the investment process.
We estimate the financing rate implicit in equity index futures (“FIR”) by comparing the prices of the near and next contracts and adjusting for expected dividends and convexity. We provide a direct estimate of the FIR volatility, along with the correlation of the FIR and the underlying stock index, which are required for the convexity adjustment and the specification of confidence intervals. Our estimates do not rely on an assumption that the FIR is similar to interest rates observed in the market. Although the volatility levels of the FIR and of market interest rates were generally comparable over our observation period, their relationship was unstable.
An empirical study over the nearly eighteen years from 1996 to August, 2013 of the spread between the FIR associated with S&P 500 futures and market interest rates was consistent with four distinct regimes, determined by the passage of major derivatives legislation in 2000, and the periods during and after the financial crisis.