The 30-second answer
Rebalancing is the act of trimming positions that have grown above their target weight and adding to positions that have shrunk below it.
Imagine you start with a portfolio of 20 stocks at 5% each. Six months later, one stock has doubled (now 9.5% of your portfolio) and another has fallen 30% (now 3.4%). To rebalance, you'd sell some of the winner and buy some of the loser, returning both back toward 5%.
That's the entire mechanic. It sounds boring. It quietly compounds into meaningful return over decades.
Why it works
The reason rebalancing adds return isn't intuitive. Most people, when they hear "sell winners, buy losers," recoil, feels like trimming the names that are working and throwing money at the names that aren't. That instinct is exactly why most retail investors don't rebalance, and exactly why rebalancing earns the alpha that it does.
The math works for two reasons:
1. Mean reversion. Stocks that have outperformed dramatically over a short period tend to underperform afterward, on average. Stocks that have underperformed dramatically tend to bounce back. This isn't a perfect rule, but it's true often enough across decades of data that systematically trimming winners and adding to losers has historically added a small but persistent amount of return.
2. Forced discipline. Rebalancing forces you to take action mechanically rather than emotionally. The investor who sells some of their best position when it's up 100% locks in a real gain. The investor who buys more of their worst position when it's down 30% gets a lower cost basis on a name they presumably believed in when it was 30% higher.
Both effects compound over time. Academic research has measured the rebalancing premium at somewhere between 0.3% and 1.5% per year, depending on the universe and rebalancing frequency. Small in any given year. Big over 30 years.
How often to rebalance
There's no single right answer, but the academic literature converges on a clear range.
Too frequent (weekly, daily). Bad. Trading costs and tax friction eat the rebalancing premium. The price moves you're reacting to are mostly noise.
Quarterly. Good. Aligns with how earnings, holdings disclosures (13Fs), and most institutional research cycles work. Frequent enough to meaningfully respond to changes, infrequent enough to avoid noise.
Annually. Acceptable but loses some benefit. By the time you rebalance, the imbalances may have grown large.
By drift threshold (e.g., when any holding hits ±25% of its target weight). A more rules-based approach. Avoids unnecessary trades but can miss opportunities if you're not paying attention.
Most professional asset managers rebalance somewhere between quarterly and semi-annually. We chose quarterly for our portfolios because:
- It aligns with the 13F reporting schedule, so when fresh data comes in, the rebalance can incorporate it.
- It's frequent enough to capture meaningful holdings changes.
- It's infrequent enough that trading costs stay small and tax considerations are manageable.
The boring example
Let's run through a clean example to make the math concrete.
You start with a $10,000 portfolio, equal-weighted across 4 stocks at $2,500 each. Stock A, B, C, D.
After six months:
- A: up 50% → $3,750
- B: up 10% → $2,750
- C: down 5% → $2,375
- D: down 20% → $2,000
Total: $10,875. The portfolio is up 8.75%.
Without rebalancing, your weights are now:
- A: 34.5%
- B: 25.3%
- C: 21.8%
- D: 18.4%
Stock A is now your biggest position by a lot, even though you didn't choose to overweight it. This is "drift", the portfolio looks different from what you intended.
With rebalancing, you reset everyone back to 25% ($2,719 each). You sell some A and B, buy some C and D.
Now suppose, over the next six months:
- A: down 20% → was overweight, now back to mean
- B: flat
- C: up 15% → reverting from oversold
- D: up 25% → reverting from oversold
The rebalanced portfolio benefits more than the unrebalanced version because you sold A near its peak and bought C and D near their lows. Across many such cycles, the small benefits add up.
This is a stylized example, real markets aren't this clean, but the directional effect has been documented across decades of empirical data.
The trap: rebalancing too aggressively
There's an opposing argument that says don't rebalance at all, or rebalance very infrequently, because trimming your winners can mean exiting truly transformative compounders too early.
If you'd been rebalancing your Apple position back to 1% every month from 2003 to 2023, you would have left an enormous amount of money on the table. Apple was a once-in-a-generation winner. Trimming it mechanically would have hurt.
This is why most rebalancing strategies use bands rather than rigid resets. Instead of "rebalance to exactly 5%," the rule might be "rebalance only if any holding drifts more than ±2 percentage points from target." That gives true compounders room to run while still preventing extreme imbalances.
We use bands implicitly. When we rebalance Core 20 quarterly, we're typically not trimming a 6% position back to 5%, too small to matter. We're trimming positions that have grown to 7-8% or more, and adding to positions that have fallen below 3%.
Tax considerations
In a taxable account, rebalancing creates capital gains. That's a real cost.
Some practical adjustments:
Use new contributions to rebalance. If you're adding $1,000/month, route the new money to whichever holdings are below target. This rebalances without selling anything.
Rebalance in tax-advantaged accounts first. IRAs and 401(k)s have no tax cost on trades. Do all your rebalancing there if possible.
Mind the holding period. Holdings under one year incur short-term capital gains rates. If you can wait a few weeks for a position to clear the one-year mark, the tax savings are usually worth it.
Tax-loss harvest opportunistically. When you do sell a losing position to rebalance, the loss can offset other gains and reduce your tax bill.
What happens at Advising Alpha rebalances
For Pro members, every rebalance comes with:
- The full updated holdings list
- The trade list (sells, then buys, with weights)
- A short note on what changed and why
- Email notification within an hour of the change
You then mirror the trades in your own brokerage account on your own schedule. Most members do it within a day or two of the email; some do it weekly or monthly. Doesn't matter much over the long run, as long as you eventually rebalance.
The full track record of every rebalance we've recorded is on the public track record page, so members and prospective members alike can see the full history.
What this means for you
If you build your own portfolio:
- Pick a rebalancing frequency (quarterly or semi-annually is best for most)
- Set a drift band (±2-3 percentage points works well)
- Use new contributions to rebalance where possible
- Do most rebalancing in tax-advantaged accounts
- Don't rebalance just because you "should", only when meaningful drift has happened
If you follow our portfolios:
- Watch for the rebalance email
- Mirror the trades within a few days
- Don't worry if you're a few days off, the model trades are a target, not a deadline
Rebalancing is one of those quiet, unsexy practices that adds real value over decades. It's the closest thing investing has to a free lunch.