Graduation Assignment Academic Excellence Program

Job Description

Your role

In the world of active portfolio management, two questions sit at the heart of every decision: what

should we own? and how much of it? The first is about direction; spotting opportunities, forming

views, deciding whether a particular asset is likely to go up or down. The second is about sizing,

deciding how much conviction to put behind each view given the risks involved. Get the direction

right and the sizing wrong, and a good idea turns into a small profit or, worse, a large loss. Get

both right, and even modest edges compound into meaningful performance over time.

Foreign exchange is the largest financial market in the world, and its prices are driven by a mix of

interest rates, capital flows, politics, and critically macro-economic news. A CPI release, a central

bank decision, or a jobs report can move a major currency pair more in ten minutes than it

typically moves in a few days, making these moments both the biggest source of opportunity and

the biggest source of risk. Many of them are also scheduled: we know weeks in advance when the

Federal Reserve will meet or when U.S. inflation data will drop, even if the outcome itself is

unknown. Yet most standard portfolio construction methods quietly ignore this, treating every day

as statistically similar to the recent past, when the calendar itself is telling us that some days are

fundamentally different. This assignment is about closing that gap.

Research assignment

For Dutch investment managers serving pension funds, foreign exchange is a critical part of

portfolio management. FX positions are used both to hedge structural currency exposures and,

increasingly, to generate returns in their own right. To do this systematically, most desks rely on

some form of mean-variance optimization (MVO); a framework that dates back to Markowitz and

forms the backbone of modern quantitative portfolio construction.

MVO takes two ingredients and turns them into a portfolio. In our case, a portfolio of currencies.

The first ingredient is a view on expected future returns. In practice, and departing slightly from

the original Markowitz formulation, these are the alphas. The second ingredient is a view on risk:

how volatile is each currency pair, and how do they co-move with one another? Given those two

inputs, MVO produces an "optimal" portfolio: one that maximises expected return for a given level

of risk, or equivalently, one that sizes each position so that pairs offering the best return per unit of

risk get the most capital, while riskier or more correlated pairs get pulled back.

In practice, most research effort on the trading desk goes into improving the return view; finding

new signals, refining them, combining them. The risk view, by contrast, tends to be treated as

plumbing: estimated in the simplest possible way from a rolling window of past returns, plugged in,

and left alone. This is where the opportunity lies. A better risk estimate does not just make the

portfolio safer; it changes how the optimizer sizes positions on any given day, and there is one

obvious source of information that standard rolling estimators throw away; the macro-economic

calendar. Currency volatility on days with scheduled events, such as FOMC decisions, U.S. CPI,

Non-Farm Payrolls, ECB meetings, or JOLTS, is systematically higher than on quiet days, and this is

known well in advance. A risk model that ignores the calendar is leaving that information on the

table, and in turn leaving better portfolio weights, and a better risk-adjusted return, unclaimed.

The broad theme of this assignment is therefore: How can knowledge of scheduled macro-economic events be incorporated into the risk estimation and portfolio construction process for FX?

The exact research question is for the student to formulate, and part of the value of this project

lies in that choice. The student is expected to converge on their own precise question in the early

weeks of the internship, in conversation with the trading team, and to justify why that question is

the one worth answering.

For this project detailed FX spot and macro-event data are available. Bloomberg terminal & API will

also be available.

Your department

This assignment is part of the unique Academic Excellence Program (AXP). This program is a

cooperation between MN, PGB and PGGM, three leading Dutch pension fund service providers.

Students will get the opportunity to work alongside investment professionals, collaborate with other

students and learn about institutional investing. On regular AXP-days all students in the program

will come together in The Hague, Amstelveen or Zeist to share ideas, learn from each other, attend

presentations from investment teams and discuss job opportunities.