What is portfolio drift, and why does it happen?
A target allocation is a decision. Drift is what the market does to that decision while nobody is looking. Understanding it is a matter of two lines of arithmetic — and the arithmetic is the point, because drift is not a risk that might materialise. It is a certainty that begins the moment the portfolio is funded.
One year of drift
Start with €100,000 in an illustrative 60/40 — €60,000 equities, €40,000 bonds. Over the year, equities return +20% and bonds return +2%:
- Equity sleeve: 60,000 × 1.20 = €72,000
- Bond sleeve: 40,000 × 1.02 = €40,800
- Total: €112,800
New weights: 72,000 ÷ 112,800 = 63.8% equities, and 40,800 ÷ 112,800 = 36.2% bonds.
Nobody made a decision. The portfolio is now carrying nearly 4 percentage points more equity than the policy says, which — given the variance arithmetic from Unit 1 — means measurably more risk than the owner signed up for.
Ten years of drift
Now compound it. Same start, and assume illustrative steady annual returns of 8% for equities and 2% for bonds over 10 years:
- Equities: 60,000 × 1.08¹⁰ ≈ 60,000 × 2.159 ≈ €129,500
- Bonds: 40,000 × 1.02¹⁰ ≈ 40,000 × 1.219 ≈ €48,800
- Total: ≈ €178,300
New weights: 129,500 ÷ 178,300 ≈ 73% equities, 27% bonds.
A 60/40 has quietly become a 73/27. If the owner's circumstances have not changed, the portfolio is now running risk they never chose — because over a decade in which equities out-returned bonds, the sleeve that grew its own weight fastest was the volatile one.
Why drift always runs toward the winner
This is structural, not coincidental. Whichever asset performs best grows its share of the total, which makes it a larger part of the next period's return, which grows its share further. Left alone, a portfolio converges toward being dominated by whatever has done best. That is a compounding feedback loop, not a market view.
Note what that does and does not fix. The direction is always toward the best performer; whether that means more risk or less depends on which sleeve won. Run the same €100,000 60/40 through 2008 — equities −37%, aggregate bonds +5% — and the sleeves finish at €37,800 and €42,000, an equity weight of 47%. Drift that year ran toward the safer sleeve, and an owner who left it alone was carrying less equity risk into 2009 than they had chosen. Over long horizons equities have usually been the winner and drift has usually meant more risk; over any particular stretch it is a fact to check rather than assume.
The consequence: a portfolio that is never rebalanced is not a passive portfolio. It is an automated momentum strategy that concentrates into recent winners. That may or may not be a fine outcome — but it is a decision, and it should be made on purpose rather than by default.
Drift within a sleeve
The same mechanism runs inside each sleeve. An equity allocation split evenly across sectors in 2015 would, by 2025, be far heavier in technology — because technology returned the most. The stated allocation ("diversified across sectors") and the actual allocation would have diverged substantially, in the direction of concentration.
In the data
Recomputing a weight as shares held × price breaks across a corporate action. Apple split each share into four on 31 August 2020:
The traded price fell by three quarters overnight and nobody lost anything: the holder had four times the shares. A drift sheet that kept the old share count and took the new price would show the Apple position shrinking by 74% in a day. The adjusted close has the split applied across the whole history, which is why the charts in this course use it, and why the two adjusted figures show the real move of about 3%.
Try it now
- Both sleeves are below, read on adjusted closes — that line already carries dividends and coupons, so it is a total return rather than a price change, which is what a drift calculation needs. Measure the last ten years on each.
- Start an illustrative €100,000 at 60/40 and apply those returns to each sleeve. What is the equity weight at the end? How far did it drift from 60%?
- Compute the trade that would restore the target: multiply the ending total by 0.60 to get the target equity value, and subtract the actual. That number is the size of the rebalancing trade — the subject of the next lesson.