Friday, November 15, 2019 — 10:30 AM EST

More information about this seminar will be added as soon as possible.

Thursday, November 14, 2019 — 4:00 PM EST

More information about this seminar will be added as soon as possible.

Friday, November 8, 2019 — 10:30 AM EST

More information about this seminar will be added as soon as possible.

Thursday, November 7, 2019 — 4:00 PM EST

More information about this seminar will be added as soon as possible.

Thursday, October 31, 2019 — 4:00 PM EDT

More information about this seminar will be added as soon as possible.

Friday, October 25, 2019 — 10:30 AM EDT

More information about this seminar will be added as soon as possible.

Friday, October 18, 2019 — 8:00 AM to Saturday, October 19, 2019 — 5:00 PM EDT
First student conference in Statistics, Actuarial Science, and Finance
Wednesday, October 16, 2019 (all day)

Xiao-Li Meng

Harvard University

Date: Thursday October 16, 2019

Title: TBD

Location: TBD

Friday, October 11, 2019 — 10:30 AM EDT

More information about this seminar will be added as soon as possible.

Thursday, October 10, 2019 — 4:00 PM EDT

More information about this seminar will be added as soon as possible.

Thursday, October 3, 2019 — 4:00 PM EDT

More information about this seminar will be added as soon as possible.

Thursday, September 26, 2019 (all day)

Optimal Transport, Entropy, and Risk Measures on Wiener space

We discuss the interplay between entropy, large deviations, and optimal couplings on Wiener space.

In particular we prove a new rescaled version of Talagrand’s transport inequality. As an application, we consider rescaled versions of the entropic risk measure which are sensitive to risks in the fine structure of Brownian paths. 

Thursday, September 19, 2019 — 4:00 PM EDT

More information about this seminar will be added as soon as possible.

Friday, September 13, 2019 — 10:30 AM EDT

More information about this seminar will be added as soon as possible.

Thursday, September 12, 2019 — 4:00 PM EDT

Nonparametric failure time with Bayesian Additive Regression Trees


Bayesian Additive Regression Trees (BART) is a nonparametric machine learning method for continuous, dichotomous, categorical and time-to-event outcomes.  However, survival analysis with BART currently presents some challenges.  Two current approaches each have their pros and cons.  Our discrete time approach is free of precarious  restrictive assumptions such as proportional hazards and Accelerated Failure Time (AFT), but it becomes increasingly computationally demanding as the sample size increases.  Alternatively, a Dirichlet Process Mixture approach is computationally friendly, but it suffers from the AFT assumption.  Therefore, we propose to further nonparametrically enhance this latter approach via heteroskedastic BART which will remove the restrictive AFT assumption while maintaining its desirable computational properties.

Thursday, August 22, 2019 — 4:00 PM EDT

Development and Application of A Measure of Prediction Accuracy for Binary and Censored Time to Event Data

Clinical preventive care often uses risk scores to screen population for high risk patients for targeted intervention. Typically the prevalence is low, meaning extremely unbalanced classes. Positive predictive value and true positive fraction have been recognized as relevant metrics in this imbalanced setting. However, for commonly used continuous or ordinal risk scores, these measures require a subjective cut-off threshold value to dichotomize and predict class membership. In this talk, I describe a summary index of positive predictive value (AP) for binary and event time outcome data. Similar to the widely used AUC, AP is rank based and a semi-proper scoring rule. We also study the behavior of incremental values of AUC, AP and the strict proper scoring rule scaled Brier score (sBrier) when an additional risk factor Z is included. It is shown that the incremental values agreement between AP and sBrier increases as the class unbalance increases, while the agreement between AUC and sBrier decreases as class unbalance increases. Under certain configurations, the changes in AP and sBrier indicate worse prediction performance when Z is added to the risk profile, while the changes in AUC are almost always favor the addition of Z. Several real world examples are used throughout the talk to illustrate and contrast these metrics.

Tuesday, August 13, 2019 — 4:00 PM EDT

Spatial Cauchy processes with local tail dependence


We study a class of models for spatial data obtained using Cauchy convolution processes with random indicator kernel functions. We show that the resulting spatial processes have some appealing dependence properties including tail dependence at smaller distances and asymptotic independence at larger distances. We derive extreme-value limits of these processes and consider some interesting special cases. We show that estimation is feasible in high dimensions and the proposed class of models allows for a wide range of dependence structures.

Monday, July 22, 2019 — 4:00 PM EDT

Negative Marginal Option Values: The Interaction of Frictions and Option Exercise in Variable Annuities


Market frictions can affect option exercise, which in turn affects the value of a marginal option to the writer—and may even yield negative marginal option values. We demonstrate the relevance of this mechanism in the context of variable annuities with popular withdrawal guarantees, both theoretically and empirically. More precisely, we show that in the presence of income and capital gains taxation for the policyholder, adding on a common death benefit option—allowing to continue the withdrawal guarantee in case of death—changes the policy- holder’s optimal withdrawal behavior. As a consequence, the total value of the contract from the perspective of the insurer may decrease, i.e. the marginal option value is negative. This explains the common practice of including death benefit options without additional charges in these products.

Thursday, July 11, 2019 — 4:00 PM EDT

On making valid inferences by combining data from multiple sources: An appraisal


National statistical agencies have long been using probability samples from multiple sources in conjunction with census and administrative data to make valid and efficient inferences on population parameters of interest, leading to reliable official statistics. This topic has received a lot of attention more recently in the context of decreasing response rates from probability samples and availability of data from non-probability samples and in particular “big data”. In this talk, I will discuss some methods, based on models for the non-probability samples, which could lead to useful inferences when combined with probability samples observing only auxiliary variables related to a variable of interest.  I will also explain how big data may be used as predictors in small area estimation and comment on using non-probability samples to produce “real time” official statistics.

Monday, June 24, 2019 — 4:00 PM EDT

A regularization approach to the dynamic panel data model estimation

In a dynamic panel data model, the number of moment conditions may be very large even if the time dimension is moderately large. Even though the use of many moment conditions improves the asymptotic efficiency, the inclusion of an excessive number of moment conditions increases the bias in finite samples. An immediate consequence of a large number of instruments is a large dimensional covariance matrix of the instruments. As a consequence, the condition number (the largest eingenvalue divided by the smallest one) is very high especially when the autoregressive parameter is close to unity. Inverting covariance matrix of instruments with high condition number can badly impact the properties of the estimators. This paper proposes a regularization approach to the estimation of such models using three regularization schemes based on three different ways of inverting the covariance matrix of the instruments. Under double asymptotic, we show that our regularized estimators are consistent and asymptotically normal. These regularization schemes involve a regularization or smoothing parameter so that we derive a data driven selection of this regularization parameter based on an approximation of the Mean Square Error and show its optimality. The simulations confirm that regularization improves the properties of the usual GMM estimator. As empirical application, we investigate the effect of financial development on economic growth. Regularization corrects the bias of the usual GMM estimator which seems to underestimate the financial development - economic growth effect.

Friday, June 14, 2019 — 10:30 AM EDT

Aggregate Risk and Bank Regulation in General Equilibrium


We examine the optimal design of bank regulation in a general equilibrium model. The unregulated economy has multiple equilibria that feature varying sizes of the financial sector and bank fragility. The economy underinvests (overinvests) in risky production when aggregate risk is low (high). We characterize and implement the efficient allocations via capital and reserve requirements, deposit insurance and bailouts. There is a range of efficient regulatory policies with a stricter capital requirement on banks being accompanied by a looser reserve requirement and less deposit insurance. We derive novel insights into how aggregate risk influences capital and reserve requirements as well as the efficiency of depositor subsidies.

Friday, May 31, 2019 — 6:00 PM EDT
MACTSC 10th Anniversary banner

In 2019, the Master of Actuarial Science (MActSc) professional degree program will be celebrating 10 wonderful years at the University of Waterloo. 

Friday, May 31, 2019 — 10:30 AM EDT
Department Logo

Worst-case risk measures and distributionally robust optimization

Distributional ambiguity refers to the situation where the probability distribution of uncertain outcomes is unknown. The question of how to account for distributional ambiguity has been of central interest in risk management, and more generally, many fields involving decision making under uncertainty. In this talk, we present a general framework of risk minimization based on distortion risk measures (also known as dual utility) and show how the worst-case risk can be evaluated when only the support and moments are known for the underlying distribution. We also show that the problem of minimizing the worst-case risk, also known as distributionally robust optimization (DRO) problem, can be solved efficiently in large scale for a large class of decision problems including portfolio optimization, production and transportation planning, among many others. Worst-case distributions, i.e. distributions attaining the worst-case risk, are characterized, which offer useful intuition about the worst-case scenarios. 

Wednesday, May 15, 2019 — 9:30 AM to Thursday, May 16, 2019 — 3:30 PM EDT

This workshop provides a crash course on using statistical methods and software when conducting data analysis in survey research. There will be a hands-on opportunity to conduct basic data analysis using SAS software. This workshop is presented by Dr. Christian Boudreau, Co-director of the Survey Research Centre (SRC), along with Grace Li from the International Tobacco Control (ITC) Project.

Friday, May 10, 2019 — 10:30 AM EDT

SURPLUS-INVARIANT RISK MEASURES ON ROBUST MODEL SPACES


In this talk, we present a systematic study of the notion of surplus invariance. In essence, the property of surplus invariance stipulates that whether or not a financial institution is adequately capitalized from a regulatory perspective should not depend the surplus profile of the company but only on its default profile. Besides providing a unifying perspective on the existing literature, we establish a variety of new results including dual representations and extensions of surplus-invariant risk measures and structural results for surplus-invariant acceptance sets. The power of our results is demonstrated in model spaces with a dominating probability, including Orlicz spaces, as well as in robust model spaces where a dominating probability does not exist.

Pages

S M T W T F S
28
29
30
31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
  1. 2019 (58)
    1. November (4)
    2. October (7)
    3. September (4)
    4. August (2)
    5. July (2)
    6. June (2)
    7. May (7)
    8. April (7)
    9. March (6)
    10. February (4)
    11. January (13)
  2. 2018 (44)
    1. November (6)
    2. October (6)
    3. September (4)
    4. August (3)
    5. July (2)
    6. June (1)
    7. May (4)
    8. April (2)
    9. March (4)
    10. February (2)
    11. January (10)
  3. 2017 (55)
  4. 2016 (44)
  5. 2015 (38)
  6. 2014 (44)
  7. 2013 (46)
  8. 2012 (44)