第103回研究会

開催日：

2025年10月17日(金) 16:00-17:00

会場：

丸の内永楽ビルディング18階 東京都立大学 丸の内サテライトキャンパス
(Zoomでも同時配信します)

報告者：

Jan Vecer 氏 (Charles University in Prague)

タイトル：

Uninformative Portfolio Choice

概要Overview

This paper proposes a new framework for portfolio optimization by representing asset prices as scaled likelihood ratios of state price densities. Within this framework, the optimal portfolio corresponds directly to the likelihood ratio between the investor’s physical measure and the market’s risk–neutral measure. In contrast to traditional approaches, which require a parametric specification of the physical measure—typically through drift $\mu$ and volatility $\sigma$—the proposed method remains agnostic about drift. The drift parameter is notoriously difficult to estimate and, in real markets, may be transient. As a result, methods that rely on its specification, such as Merton’s portfolio problem, can fail, sometimes spectacularly. By contrast, the optimal measure here is identified ex post from observed data, linking portfolio choice to maximum likelihood estimation and, in the large-sample limit, to Bayesian model selection. This formulation provides a non-parametric route to extracting drift and positions portfolio optimization within a statistical model–selection framework.

講師ご略歴Profile for speakers

Jan Vecer is a full-time faculty member at Charles University in Prague. He received his Ph.D. in Mathematical Finance from Carnegie Mellon University in 2000. From 2001 to 2010, he was on the faculty of the Department of Statistics at Columbia University, before returning to Europe to join the Frankfurt School of Finance and Management, where he was a full professor of finance from 2010 to 2015 (and has since remained affiliated as a visiting professor). In addition to his academic work, he has served as a consultant, developing market-making trading bots in major prediction and energy markets. He is the author of Stochastic Finance: A Numeraire Approach (CRC Press, 2011) and is currently completing Principles of Portfolio Choice, forthcoming in 2026.

お申込み方法Registration

第103回研究会へ参加される方は必ず事前登録をお願いいたします。
参加登録の締切は10月15日(水)です。
下記のいずれかの方法で「お名前・所属機関・電子メールアドレス」をご連絡ください。
開催日当日の午前までにご参加のために必要なURLを記載したメールをお送りします。
※参加費は無料です。

WEBからの登録
本ページにあります「REGISTER HERE」ボタンをクリックいただき、お申し込みフォームよりご登録ください。

電子メールでの登録
finance●tmu.ac.jpまで「お名前・所属機関・電子メールアドレス」をご連絡ください。
（※ご連絡時は●を@に置き換えてご使用下さい。）

ご参加情報は、東京都立大学金融工学研究センター、講演者により共有され、当セミナーの円滑な運営のために利用します。