A written factor-tilt policy: allocation, underperformance tolerance in years and review rules

What you will learn

• The sections a factor-tilt policy needs and how it differs from the general investment policy.
• How to size the tilt from a tracking-error budget and translate it into euro amounts.
• How to write an underperformance tolerance in years and in percentage points using the formula from lesson 11.
• Review rules that separate "the plan is working as expected" from "the plan needs changing".

Explanation

Why a separate document

The course "ETF Investing" ended with a one-page investment policy. A factor tilt needs an extra page because it adds relative risk against the market and a new temptation: abandoning the tilt at the worst time. The policy is written while you are calm and read when you are not; keep it to one page and specific enough that a future you cannot argue with it.

Section 1 – purpose and expected outcome

One paragraph: which factors, why they may continue (the four tests of lesson 11), and an honest expected net premium from the pessimistic-to-moderate range of lesson 10, never the academic gross figure. Write: "This is a historical observation, not a forecast; the tilt may underperform for more than ten years."

Section 2 – allocation and tracking-error budget

Decide the tracking error you accept for the whole portfolio against your market-cap benchmark; for most private investors 2–4 % a year is already demanding. If the sleeve has TE_s against the market and the rest is the market itself, portfolio TE ≈ sleeve weight × TE_s. State the weights in percent and euros, the products, and the maximum weight per factor and per ETF. The next course applies the same budget idea to satellites of any kind.

Section 3 – underperformance tolerance

The heart of the document. Write three things: the years of lagging you will tolerate without action (the patience horizon); the cumulative shortfall in percentage points at which you will review (not sell); and the conditions for an actual exit. A useful review trigger is two tracking-error standard deviations below the expected path: trigger(N) = μ × N − 2 × σ × √N. Crossing it means bad luck beyond the 2.5 % tail or a wrong assumption; both deserve a structured review, neither an automatic sale.

Section 4 – review and change rules

Annual review on a fixed date: recompute loadings (lesson 9), check turnover and tracking difference of each product (lesson 10), compare realised shortfall with the trigger path, and rebalance if any sleeve has drifted beyond its band. Policy changes require a written reason, a waiting period (for example 90 days) and must not be made in the month of a loss-triggered review. Exits only for stated reasons: the economic rationale has failed, the product changed its index, costs exceed a stated ceiling, or your risk capacity has changed.

Worked example

Hypothetical example: portfolio of €100,000 in equities, benchmark a world market-cap ETF. The investor accepts a portfolio tracking error of about 1.2 % and chooses an integrated multi-factor sleeve with TE 4 % against the market, expected net premium 1.5 % on the sleeve. Taxes and dealing costs are ignored.

Step 1 – sleeve size: 1.2 % / 4 % = 30 %, so €30,000 tilt and €70,000 core in the market-cap ETF.

Step 2 – split of the sleeve: value €15,000, momentum €7,500, quality €7,500 (15 % / 7.5 % / 7.5 % of the portfolio); maximum per single ETF 15 %.

Step 3 – expected contribution: 30 % × 1.5 % = 0.45 % a year on the whole portfolio, about €450 a year in expectation. Writing it down prevents disappointment: a modest improvement, not a transformation.

Step 4 – tolerance for the sleeve (μ = 1.5 %, σ = 4 %): trigger(N) = 1.5 % × N − 2 × 4 % × √N.

The investor writes: "Patience horizon 10 years. Review if the sleeve trails the benchmark by more than 10 percentage points cumulatively at any annual review. No exit for performance reasons alone before year 10."

Step 5 – rebalancing band: each sleeve ±25 % of target (value 15 % → 11.25 % to 18.75 %). After a bad year the value sleeve is €12,000 of a €95,000 portfolio, 12.6 %: inside the band, no trade.

Common mistake

Practice activity

On paper: you have €60,000 in equities, accept a portfolio tracking error of 1.5 %, and want a single small-cap value ETF (TE 6 % against the market, assumed net premium 2 %). (a) Size the sleeve in percent and euros. (b) Compute the expected annual contribution in euros on the whole portfolio. (c) Compute the five-year review trigger for the sleeve. (d) Write one exit rule that is not performance-based.

Key takeaways

• A factor-tilt policy adds purpose, allocation, tolerance and review rules to your general investment policy, on one page.
• Size the tilt from a tracking-error budget: sleeve weight ≈ portfolio TE / sleeve TE, then write it in euros.
• Pre-commit a patience horizon in years and a review trigger of μN − 2σ√N; reviews, not automatic sales.
• Allow exits only for stated non-performance reasons, after a waiting period.

Next step

You now have a complete, evidence-based factor plan; continue with the course "Portfolio Construction in Depth: Risk Budgets, Regimes and Stress Tests", which places this tilt inside a full risk budget with regimes and stress tests.