Quant Matters – Risk-Reward Revisited: A Conviction-Driven Alpha Framework

研报英文原文证据摘录

Quant Matters – Risk-Reward Revisited: A Conviction-Driven Alpha Framework

Global IdeaMdirection. Short signals targeted the converse – contrarian downward revisions in which

the stock had yet to sell off and the downside-adjusted tilt was already punitive, capturing

instances in which the analyst's conviction had quietly shifted before the market had

reflected it.

In this edition of Quant Matters, we revisit the Risk-Reward strategy with updated data to

29 May 2026. Nearly six years have passed since the original publication, providing a

meaningful out-of-sample window to reassess whether analyst-driven alpha, expressed

through Risk-Reward, remains durable, differentiated and economically relevant. The

remainder of the report proceeds in three steps. First, we review the Risk-Reward

framework and the variables that translate analyst scenarios into systematic signals.

Second, we evaluate the updated performance record. Third, we test whether the Risk-

Reward filters add value beyond simpler base-case revision strategies, and whether the

signal survives a more practical monthly implementation.

Exhibit 1: Illustration of the Risk-Reward Framework

Source: Morgan Stanley Research

Revisiting the Risk-Reward framework

At the core of the Risk‑Reward framework are three 12‑month analyst price forecasts: the

bull case, the base‑case price forecast, and the bear case. These represent optimistic,

central and pessimistic scenarios for the stock. In the implementation, we use the

base‑case price forecast as the central estimate rather than the displayed price target, so

as to explicitly focus on analysts' evaluation of three cases. We note that the base-case

forecasts and price targets coincide in most cases, hence this choice does not materially

affect the results.