Global Equity Portfolio Construction

University of Edinburgh · Investment Management · March 2026

End-to-end design of a GBP 10 million long-only global equity fund: multi-factor selection, constrained optimisation, ESG integration, and institutional risk reporting.

Snapshot

£10M

AUM

Long-only global equity

~29

Holdings

9 sectors · 14+ countries

6.8%

Model TE

Process cap 8%

7%

Max name

Sector cap 35%

Benchmark	MSCI World (iShares IWRD LN)
DM / EM	~80% / 20%
ESG	Name-level Bloomberg floor; portfolio top-quartile of universe

Process

1

Mandate

MSCI World, TE budget, ESG floor, DM/EM mix

2

Score

13-factor DM/EM composites, sector-neutral

3

Select

Liquidity · free float · veto · greedy caps

4

Optimise

Factor cov · MV / CVaR / robust (CVXPY)

5

Risk

TE · VaR suite · stress · FX

Factor model

Security selection

DM composite factor weights

Quality and momentum lead; valuation and growth balance the book; secondary tilts include efficiency, conviction, short interest, and dividend.

Portfolio structure

Allocation

Sector weights

Active IT sleeve with industrials and financials diversification.

Active risk

Tracking-error decomposition

Idiosyncratic + style dominate; market contribution kept small.

Mandate constraints

Constraint	Limit
Single name	≤ 7%
Sector	≤ 35%
Sub-industry	≤ 15%
Tracking error	≤ 8% process / 15% mandate
Liquidity	Min market cap & ADV
ESG	Minimum Bloomberg ESG per holding

Risk analytics

Loss estimates

VaR / CVaR at 95% (1-month, % of NAV)

Parametric and Cornish–Fisher sit near ~10% of NAV; historical VaR is lower — both families are reported.

Stack

Python · pandas · NumPy · SciPy · CVXPY · scikit-learn · Bloomberg equity/ESG · config-driven pipeline

Competencies

Investment process · multi-factor selection · constrained optimisation · factor risk · VaR/CVaR · ESG integration

---

← All projects