Portfolios drift toward whatever recently worked, and nothing stops that quietly accumulating risk while every holding still looks sensible individually. Left alone, a plan built around balanced exposure becomes increasingly dependent on the categories having good fortune simultaneously – which is precisely the condition nobody chose consciously. Rebalancing is the correction, and its reputation as tedious bookkeeping understates how much of long-run experience it determines.
Mechanics with labelled round numbers so nothing requires trust. Take $10,000 split six-tenths growth assets and four-tenths defensive assets, both performing as their labels suggest over several years with the stronger category compounding uninterrupted: weights migrate, defensively shrinking, so what began as balance ends concentrated in whatever enjoyed good fortune while nobody re-examined anything. Each intervening year looked like confirmation rather than warning, since recent strength feels like evidence about quality rather than luck – history’s winners relocate constantly. Restoring targets therefore means trimming the appreciated category and adding to the neglected one, which is sell-high, buy-low conducted mechanically rather than through judgment no human performs comfortably while momentum flatters the alternative.
Bands beat calendars for most people. Checking quarterly regardless of conditions forces transactions from ordinary noise, generating costs without addressing meaningful drift. Acting only when a category strays several percentage points from target captures genuine imbalance while ignoring irrelevant fluctuation. Research comparing approaches reaches ambiguous conclusions about which produces better arithmetic, depending heavily on sample period and whether trading expenses were included honestly – anyone claiming certainty about optimal frequency has read one study. The choice matters far less than consistently executing some rule, because the failure mode is not suboptimal scheduling but abandoning discipline whenever markets feel exciting.
Contribution-based adjustment avoids selling entirely. New money directed at underweight categories moves weights toward targets without realising gains, which matters substantially inside taxable accounts where sales generate filings and possibly bills – keeping more of your gains involves structure as well as choices. During accumulation years this handles most drift; only once withdrawals begin do sales become unavoidable, at which point distribution sequencing should deliberately draw from overweight positions rather than whatever seems convenient – withdrawal order changes outcomes dramatically.
Volatility harvesting needs realistic framing. Combining categories that move independently permits periodic restoration of weights to capture relative movements, sometimes improving returns for comparable risk in simulations. Three caveats deserve equal prominence: it requires genuinely unsynchronised components, correlations converge precisely when separation gets needed – cushion evaporates exactly when required; it presumes continued existence of the diversifying asset rather than its disappearance under stress; and simulations routinely exclude bid-ask friction and account restrictions applicable to small balances. Treat as a reasonable practice rather than a discovered free lunch.
Costs, taxes, and the trap nobody warns about
Sales create expenses, spreads and tax events whose significance scales inversely with portfolio size – on modest balances a fixed charge consumes multiple years of marginal benefit, favouring fewer adjustments executed deliberately over frequent tinkering. Every transaction also represents an implicit forecast that current relative prices are wrong; if you hold strong convictions about direction, adjusting contradicts them, which is fine but should be conscious rather than accidental.
The trap worth naming: investors frequently rebalance into something else instead – rotating from lagging category into whichever category currently demonstrates superior results. That isn’t restoring targets; it’s momentum chasing wearing disciplined vocabulary, and it converts a genuinely useful mechanical process into market timing with extra paperwork. Rules distinguish the two immediately: target weights written in advance and rarely revised versus targets quietly updated following recent performance. If the numbers move toward attention rather than away from drift, the exercise has inverted into its opposite, which explains why plenty of diligent traders feel organised while steadily increasing concentration in whatever they’re currently enjoying.
Sequence deserves emphasis near any date when cash must arrive. A portfolio drifting heavily into growth assets entering a period of required withdrawals faces adverse ordering effects that averages conceal completely; restoring defensive weight before deadlines begin removes dependence on favourable timing that nobody controls – order of returns outweighs average return here. For younger investors the same logic applies across goals rather than across days: money destined for a vehicle within two years belonging predominantly in growth assets indicates planning confusion unrelated to rebalancing frequency.
Practical implementation for someone starting with a regular surplus: write targets down somewhere durable including the reasoning, since the document becomes valuable during the first decline making everything feel obviously mistaken – the moment when most actual damage occurs and professionals remain equally susceptible. Check weights annually rather than weekly, since intraday monitoring converts noise into emotional obligations and no advantage accompanies daily observation of numbers moving randomly. Adjust using contributions primarily; execute full sales only when bands breach meaningfully. Record decisions briefly with dates, because memory retroactively edits confidence and later review of what was actually anticipated builds better judgement than any strategy detail. And resist sophistication creep: adding categories beyond capacity to monitor creates maintenance burden plus false precision, while two or three broad exposures covering most of an economy handle the underlying problem adequately for decades. Whatever structure you choose remains subordinate to whether you can continue owning it through unpleasant stretches, which measures portfolio quality better than any optimisation ever attempted on the way in.
This is educational explanation of portfolio maintenance, not financial advice or a recommendation of any product, allocation or account type; all figures are invented round numbers supplied so arithmetic can be checked independently, published findings depend on measurement windows and cost assumptions, tax treatment varies by wrapper and jurisdiction, rules change with legislation, investing involves risk, and rebalancing cannot prevent loss or guarantee improvement over simply holding.