Asset Allocation Explained: Splitting a Portfolio

Almost every argument about investing is really an argument about something small. Which fund, which company, which day to buy, whether the expense ratio should be 0.03 percent or 0.09 percent. Those questions have answers and the answers matter at the margin, but they sit downstream of a decision that most people make casually and then never look at again: what fraction of the money goes into stocks, what fraction into bonds, and what fraction sits in cash. That split is the asset allocation, and it explains far more about how a portfolio behaves than the contents of any single sleeve.

This explainer takes the split apart. It covers what each asset class is actually for, how time horizon and the ability to absorb a loss set the mix, why the popular age rules disagree with each other, how glide paths work, where bonds fit and why they behave differently when rates are rising than when they are falling, how international and sector tilts change the picture, why cash is a sleeve and not a strategy, the difference between asset allocation and asset location, and what happens to a mix that nobody ever checks. It sits alongside our work on what portfolio management is, how to rebalance your portfolio and how bonds work, and it leans on those rather than repeating them. Every figure below is invented round arithmetic chosen for internal consistency, not a forecast and not a description of any real market.

What asset allocation actually decides

Asset allocation is a set of percentages that add to 100. It says how much of the total sits in each broad category of investment, and that is the whole of it. It says nothing about which fund, which company, which country or which manager, and it can be written on the back of an envelope in about fifteen seconds.

What makes such a simple statement powerful is that the categories behave differently from each other. Stocks and bonds do not rise and fall in the same rhythm or by the same amounts. A portfolio that is 90 percent stocks and one that is 30 percent stocks are not two versions of the same thing with a dial turned slightly; they are different experiences, particularly in the years nobody enjoys.

Because of that, the allocation determines the shape of the ride. It sets roughly how much growth you are signing up for, roughly how much decline you are agreeing to sit through, and roughly how much of the outcome depends on markets rather than on your own contributions. Everything else you choose operates inside that shape.

The practical consequence is an ordering. Decide the split first, then decide how to fill each sleeve. Choosing funds before choosing an allocation is how people end up owning nine holdings that are all the same bet, which is a problem no amount of fund research fixes.

Allocation, diversification, and selection are three different jobs

These three words get used interchangeably and they are not the same job. Allocation is the split across asset classes. Diversification is the spreading of risk inside a sleeve so that no single holding decides the outcome. Selection is the choice of the individual holdings themselves.

You can do one well and the others badly. A portfolio holding forty different stocks is beautifully diversified inside its equity sleeve and completely undiversified across asset classes if stocks are all it holds. A portfolio holding one broad index fund and one broad bond fund has almost no selection work in it at all and yet is allocated and diversified perfectly adequately.

The reason to keep them separate in your head is that they fail differently. Bad selection costs you the difference between one holding and another, which is usually small if you are holding broad funds. Bad diversification costs you when one holding fails, which is occasional and survivable if the position was modest. Bad allocation costs you in every year the market moves, which is every year.

Selection work is company-level work, which is why our walkthroughs on how to read an earnings report and what stock buybacks do belong to that layer rather than this one. Our piece on how to invest in index funds covers the filling of a sleeve, and index funds versus ETFs covers the wrapper question. This explainer stays on the split.

The major asset classes and what each is actually for

Most portfolios are built from three or four broad categories, and each is included for a specific job rather than because a list said so. Naming the job is useful, because a holding that is not doing its job is easy to spot once you know what you asked it to do.

Stocks are the growth engine. They represent ownership of businesses, they carry the widest range of outcomes, and over long stretches they are the reason a portfolio outpaces the cost of living. They are also the source of nearly all the pain.

Bonds are the ballast. They represent lending rather than owning, they pay a contractual stream, and they are held to reduce the size of the fall in the years stocks fall. They are not held to make you rich.

Cash is the buffer. It is held so that a spending need or an emergency does not force a sale at whatever price the market happens to be offering that week. It is measured in months of spending, not in percentages of ambition.

Real assets, including property exposure through listed vehicles, sit somewhere between. They are held for a return stream driven by rents and physical assets rather than by corporate earnings alone, and they are usually a minority position. Our note on what a REIT is covers that sleeve in detail.

Stocks: the growth engine and the source of the drawdown

The stock sleeve is where the expected return comes from and where the volatility comes from, and those two facts are the same fact. You are compensated over long periods precisely because the short periods are uncomfortable enough that many people do not stay invested through them.

This explainer uses an illustrative long-run assumption of 7 percent a year for stocks and an illustrative bad-year decline of 40 percent. Both numbers are invented round figures chosen to make the arithmetic visible. Real markets do not deliver a smooth 7 percent and do not confine their declines to a tidy 40 percent, and any single decade can look nothing like the assumption.

The important property of the stock sleeve is not its average but its dispersion. An asset that averages 7 percent by returning 25 percent one year and losing 15 percent the next is a completely different thing to live with than an asset that returns 7 percent every year, even though a spreadsheet may report the same average. Our note on what CAGR is works through why the average of the annual figures and the compounded growth rate are not the same number.

Inside the sleeve, the sensible default for most people is broad ownership rather than concentrated bets, because concentration adds dispersion without reliably adding expected return. That is a selection decision, and it sits underneath the allocation decision rather than replacing it.

Bonds: the ballast that is not always calm

A bond is a loan with a schedule attached. You lend a sum, you receive interest at agreed dates, and you are repaid at maturity. That contractual structure is why the sleeve is steadier than stocks, and it is also why the sleeve is not risk free.

The two things that move a bond’s value are the direction of interest rates and the market’s view of whether the borrower will pay. A bond bought when rates were low loses market value when new bonds are issued at higher rates, because nobody pays full price for the older, stingier stream. A bond issued by a shakier borrower falls in price when confidence in that borrower weakens.

This explainer uses an illustrative bad-year decline of 8 percent for the bond sleeve and an illustrative long-run return of 4 percent. That 8 percent figure is deliberately not zero, because the common belief that bonds cannot fall is the single most expensive misunderstanding in the ballast sleeve. Bonds fall less than stocks in most bad equity years. They do not fall never, and there are periods when both sleeves fall together.

For the mechanics of why price moves opposite to rates, and what maturity does to the size of that move, our explainer on how bonds work is the fuller treatment, and what a bond ladder is covers one way to structure the sleeve around known spending dates.

Cash: a sleeve, not a strategy

Cash earns the least of the three over long periods, and holding a large permanent cash position is one of the more expensive habits available to a long-horizon investor. Yet a portfolio with no cash at all forces every unexpected expense to be funded by selling something, occasionally at the worst possible moment.

The resolution is to treat cash as a sized sleeve with a stated job rather than as a place to wait for clarity. Its job is to cover near-term spending and emergencies, so it is sized in months of expenses, and once that requirement is met the balance belongs in the growth and ballast sleeves. This explainer uses an illustrative 5 percent cash weight, which on a $400,000 portfolio is $20,000.

The failure mode worth naming is cash as a market opinion. Money moved to cash because stocks feel expensive is a timing decision wearing the clothes of prudence, and it requires two correct calls rather than one: when to leave and when to return. The second call is the one that is almost never made, because the conditions that make returning feel comfortable arrive only after the recovery.

Our note on how high-yield savings accounts work covers where the sleeve can sit, and treasury bills explained covers a short-dated alternative for the same job.

Real estate and other real assets in the mix

Listed property vehicles give a portfolio exposure to rents and physical assets without requiring anyone to own a building directly. They are usually included as a minority slice, and the argument for including them is that their return stream is driven by different forces than corporate earnings, so the sleeve does not always move in step with the stock sleeve.

The argument against a separate slice is that broad stock index funds already contain listed property companies, so a dedicated allocation is a tilt rather than a new asset class. That is a fair objection, and it means the honest way to describe a property slice is as an overweight to something you already own a little of.

Practically, a property tilt behaves much more like the stock sleeve than the bond sleeve during severe market declines, which matters when you are estimating how far the whole portfolio can fall. Treating it as a diversifier and then discovering it fell alongside stocks is a common disappointment.

The illustrative mix in this explainer keeps things to three sleeves for clarity, so any property exposure is assumed to sit inside the stock sleeve rather than beside it. That is a simplification, and it is stated rather than hidden.

The illustrative portfolio this article uses

Every number from here forward comes from one invented portfolio, so the arithmetic can be followed end to end. The portfolio is worth $400,000 and its target allocation is 60 percent stocks, 35 percent bonds and 5 percent cash.

In dollars, that is $240,000 in stocks, $140,000 in bonds and $20,000 in cash. Inside the stock sleeve, 70 percent is domestic and 30 percent is international, which works out to $168,000 and $72,000, or 42 percent and 18 percent of the whole portfolio.

The assumed long-run returns are 7 percent for stocks, 4 percent for bonds and 3 percent for cash. The assumed bad-year declines are 40 percent for stocks, 8 percent for bonds and 0 percent for cash. These are invented figures picked to be round and internally consistent, not estimates of anything.

Blending those assumptions by weight gives the portfolio’s headline numbers. The expected return is 60 percent of 7 plus 35 percent of 4 plus 5 percent of 3, which is 4.20 plus 1.40 plus 0.15, or about 5.75 percent a year. The bad-year decline is 60 percent of 40 plus 35 percent of 8, which is 24.0 plus 2.8, or about 26.8 percent. Put your own balance and target percentages into the companion below to see both figures move.

Time horizon: the input that does most of the work

Time horizon is how long the money can stay invested before you need to spend it, and it is the single most useful input into an allocation because it changes what a decline actually costs you. A 30 percent fall in money you will not touch for twenty five years is a paper event that you have time to recover from. The same fall in money you need in eighteen months is a permanent reduction in what you can buy.

Horizon is not one number for most people. A house deposit due in three years, a child’s education in twelve and a retirement in twenty eight are three horizons living in one household, and treating them as a single pot means the shortest goal ends up carrying the risk that only the longest one can afford.

The cleaner approach is to allocate per goal. Money with a short horizon leans heavily toward cash and short bonds, because certainty of amount matters more than growth. Money with a long horizon leans heavily toward stocks, because growth matters more than smoothness and there is time to absorb the bad years.

Retirement complicates this, because it is not a single date but a spending period that may last decades. That means a retirement portfolio holds both short-horizon money for the next few years of spending and long-horizon money for the twenties and thirties years out, which is one of the reasons a retiree’s allocation is rarely all bonds. Our work on how much you need to retire sets the target the whole plan is serving, and the calculator on our homepage anchors that number.

Ability to take risk versus willingness to take it

Two people with identical ages and identical balances can correctly hold different allocations, because risk has two separate components and they are frequently in conflict.

Ability to take risk is a matter of arithmetic and circumstance. It rises with a longer horizon, with a larger surplus over what you actually need, with stable income, with low fixed obligations, and with guaranteed income sources that cover the essentials regardless of what markets do. A person whose basic expenses are already covered by a pension has a high ability to take risk with the rest, whatever their age.

Willingness is a matter of temperament, and it is not improved by being told it should be higher. It reveals itself in behaviour, not in questionnaires: what you actually did the last time your balance fell by a fifth is worth more evidence than any answer to a hypothetical.

The binding constraint is the lower of the two. An allocation that your circumstances can support but your nerves cannot is worse than a more modest one, because the failure mode is selling at the bottom, and a plan abandoned in the worst month costs more than the extra growth it was designed to capture. Building a mix you will actually keep through a bad year is the whole objective, and a somewhat conservative allocation held for thirty years beats an aggressive one held for three.

What an illustrative mix costs you in a bad year

Abstract percentages are easy to agree to. Dollar amounts are not, and converting one to the other is the most useful five seconds in this entire subject.

The illustrative $400,000 portfolio at 60/35/5 carries an assumed bad-year decline of 26.8 percent. In dollars that is $107,200, leaving $292,800. That is the number to sit with, because a 26.8 percent decline is a statistic and a $107,200 decline is an experience.

Run the same exercise across the mixes in the chart above. All stocks would fall 40.0 percent, which is $160,000. An 80/20 mix would fall 33.6 percent, or $134,400. A 40/60 mix would fall 20.8 percent, or $83,200. A 20/80 mix would fall 14.4 percent, or $57,600. The difference between the most and least aggressive of those is $102,400 of paper loss in a single bad year on the same starting balance.

Set against that, the assumed return difference is 7.0 percent versus 4.6 percent a year. Over long stretches that gap compounds into a very large sum, which is precisely why the trade is a real trade rather than a free choice. You are buying a smaller bad year with a smaller expected end balance, and the right point on that line depends on how much end balance you need and how much bad year you can hold. Enter your own numbers in the companion below to see the dollar version of your own mix.

The classic heuristics and their honest limits

The best known shortcut is to hold your age in bonds, so a 40 year old holds 40 percent bonds and 60 percent stocks. A newer family subtracts your age from a fixed number to get a stock weight, most commonly 110 minus age or 120 minus age, which put the same 40 year old at 70 percent or 80 percent stocks respectively.

Notice what has happened. Three widely repeated rules give the same person 60, 70 or 80 percent stocks, a spread of twenty percentage points. On the illustrative decline assumptions that is the difference between a bad year of about 24.8 percent and one of about 32.0 percent, which is $99,200 and $128,000 on $400,000. The rules cannot all be right, and the reason they differ is that each was built for a different assumed life expectancy and a different assumed appetite.

The deeper limitation is that all of them use one input. Age is a proxy for horizon, and horizon is only one of the three things that should set the mix. None of the rules asks whether you already have enough, whether your income is stable, whether a guaranteed income covers your essentials, or what you did the last time markets fell.

Used properly, a heuristic is a starting point that you then argue with. Write down the number the rule gives you, then list the reasons your circumstances push it higher or lower, then land somewhere and record why. The written reason is the part that survives the next bad quarter.

Glide paths: allocation as a function of years left

A glide path is a rule that changes the allocation gradually as a target date approaches, rather than leaving it fixed and then changing it all at once. The typical shape starts stock heavy while the horizon is long and steps the stock weight down as the spending date nears, so the money becomes progressively less exposed to a decline it would not have time to recover from.

The virtue of a glide path is that it is a decision made once, in advance, and then executed mechanically. Deciding to reduce risk at 60 is easy to do at 45 and surprisingly hard to do at 60, when the reduction usually means selling something that has done well.

There is a second, less intuitive shape worth knowing about. Some approaches lower the stock weight into the first years of retirement and then raise it again over the following decades, on the reasoning that the period of greatest vulnerability is the handful of years either side of the spending start, not the whole retirement. Our note on sequence of returns risk explains why that window is the dangerous one.

Whichever shape you choose, the mechanics matter less than writing it down as a schedule with dates and target weights attached. A glide path that lives only as an intention tends to glide nowhere.

Why bonds behave differently in different rate environments

The bond sleeve does not have one personality. What it does for your portfolio depends heavily on where interest rates are and where they are going, and this is the part most allocation discussions skip.

When rates are falling, existing bonds become more valuable, because their fixed payments look generous next to newly issued ones. The sleeve then does two jobs at once: it pays its interest and it gains in market value, often at exactly the moment stocks are struggling. That is the behaviour people have in mind when they call bonds a hedge.

When rates are rising, the same mechanism runs in reverse. Existing bonds lose market value as new issues offer more, and the sleeve can post a negative year. If that rise in rates happens for reasons that also hurt stocks, both sleeves fall together and the diversification you were counting on is absent in the year you needed it.

There is a third state that matters for expectations. When starting yields are low, the sleeve’s future return is low almost by definition, because a bond’s return over its life is dominated by the yield you bought it at. A bond sleeve is therefore not a fixed 4 percent contributor; the illustrative 4 percent used here is an assumption, and the honest version is that the current yield on the sleeve you hold is the best available anchor for what it will deliver.

Duration: the dial that sets how a bond sleeve moves

Within the bond sleeve there is a dial that decides how much of the above you experience, and it is duration. Duration measures how sensitive a bond or bond fund is to a change in rates, and it rises with the length of time until the money comes back.

The practical shorthand is that a fund with a duration of roughly five years loses roughly 5 percent of value for each one percentage point rise in rates, and gains roughly the same for each one point fall. It is an approximation rather than a law, but it is close enough to be useful for sizing.

That single number lets you match the sleeve to its job. If the bond sleeve exists to steady a long-horizon portfolio, a longer duration gives more of the cushioning effect when rates fall during an equity decline. If it exists to hold money you will spend in three years, a long duration is the wrong tool, because the value can move more than the amount you were trying to protect.

Many portfolios end up with a bond sleeve whose duration nobody chose, inherited from whichever broad fund was convenient. That is not a disaster, but it is worth knowing the number. Choosing it deliberately is the difference between ballast and an unexamined second bet.

International exposure: the tilt everyone argues about

Once the stock sleeve is sized, the next question is how much of it goes abroad, and this is one of the genuinely unsettled arguments in portfolio construction.

The case for a meaningful international weight is concentration. A domestic-only stock sleeve is a large single bet on one economy, one currency and one policy regime. Diversifying across regions reduces the chance that a long stretch of poor domestic performance defines your entire result, and no one gets to know in advance which region will lead.

The case against is threefold. Large domestic companies already earn revenue globally, so the exposure is not as absent as the label suggests. Currency movements add volatility to returns measured in your own currency, which can swamp the underlying performance for years. And foreign holdings can carry higher expense ratios and more complicated tax treatment, a cost our note on what an expense ratio is shows compounding quietly.

The illustrative portfolio here puts 30 percent of its stock sleeve abroad, which is 18 percent of the whole. That is a middle position chosen so the arithmetic has a number in it, not a recommendation. What matters more than the specific figure is that you pick one, write down the reasoning, and then leave it alone, because a weight that changes every time one region has a bad year is not a policy.

Sector and factor tilts: small weights that feel large

Beyond the regional split, portfolios accumulate tilts: extra weight in technology, or energy, or smaller companies, or higher dividend payers. Some are deliberate and some arrive because a holding grew.

The first thing to establish about any tilt is its actual size relative to the whole portfolio. A 10 percent position in a sector fund inside a 60 percent stock sleeve is 6 percent of the portfolio, which is unlikely to change your outcome much in either direction. People routinely overestimate how much a satellite position is doing, in both directions.

The second thing is whether the tilt is what you think it is. A dividend-focused fund is often a tilt toward particular sectors as much as a tilt toward income, so a portfolio holding both a dividend fund and a utilities fund may be doubled up without realising it. Our work on how to build a dividend portfolio treats that overlap as a construction problem in its own right.

The third is the honest cost. Tilts usually carry higher fees than broad exposure and they generate more trading, and both of those are certain while the extra return is not. That does not make tilts wrong. It makes them something to size deliberately and to review against the reason you added them, rather than something to accumulate.

Asset location: which account holds what

Asset allocation asks what you own. Asset location asks which account it sits in, and the two are entirely independent decisions that are constantly confused with each other.

Start by measuring the allocation across every account combined. A tax-deferred workplace account, a Roth-style account and a taxable brokerage account are three logins and one portfolio, and an allocation calculated inside a single account can be badly wrong about your actual exposure. Someone whose workplace account is all bonds and whose brokerage account is all stocks does not hold two portfolios; they hold one, and only the combined percentages describe it.

Once the combined split is set, location is about placement. The general principle is that holdings which throw off income taxed annually are the ones that benefit most from sitting in a sheltered account, because the shelter is worth most where the tax would otherwise be charged every year. Holdings that generate little annual taxable income and are held for a long time waste less of that shelter and can sit in the taxable account.

Two cautions. The rules that make this work, including which accounts exist, how each is taxed and what the contribution limits are, differ by jurisdiction and change over time, so confirm current treatment from a primary source rather than from any article. And our comparison of a Roth IRA and a 401(k) covers the account choice itself, which is a separate question from what you put inside them.

A worked example: the target mix and one year of drift

Follow the illustrative portfolio through a single year. It starts at $400,000 split 60/35/5: $240,000 stocks, $140,000 bonds, $20,000 cash.

The year is a strong one for stocks and a flat one for bonds. Stocks return 25 percent, bonds return negative 2 percent, cash returns 3 percent. These are invented single-year figures, deliberately different from the long-run assumptions, because a single year almost never matches the average.

Sleeve by sleeve, stocks become $300,000, bonds become $137,200, and cash becomes $20,600. The portfolio total is $457,800, a gain of $57,800, or about 14.5 percent for the year.

Now recompute the percentages. Stocks are $300,000 divided by $457,800, which is about 65.5 percent. Bonds are about 30.0 percent. Cash is about 4.5 percent. Nothing was bought or sold, and yet the mix is no longer the one that was chosen: the stock weight is 5.5 percentage points above target.

That drift changed the portfolio’s risk. Blending the decline assumptions at the new weights gives about 28.6 percent rather than 26.8 percent, so a bad year would now cost about $130,900 on the new balance instead of the $122,700 the target mix would have implied. The portfolio got more aggressive without anyone deciding it should. Put your own balance, targets and stock return into the companion to run this same year on your figures.

How allocation drifts if you never look

One year of drift is small. The reason drift deserves a section is that it compounds, and it compounds in one direction, because the sleeve with the highest expected return grows fastest and therefore takes an ever larger share of a portfolio nobody touches.

Apply the long-run assumptions to the starting portfolio and leave it entirely alone. Stocks compound at 7 percent, bonds at 4 percent, cash at 3 percent. After ten years the stock sleeve is about $472,100, the bond sleeve about $207,200, and cash about $26,900, a total near $706,200. The stock weight is now about 67 percent rather than 60.

Run it another decade. After twenty years the sleeves are roughly $928,700, $306,800 and $36,100, a total near $1,271,600, and the stock weight is about 73 percent. A portfolio described to its owner as moderate has quietly become aggressive.

Translate that into risk. At 67 percent stocks the blended bad-year decline is about 29.1 percent instead of 26.8. At 73 percent it is about 31.3 percent. The portfolio is being pushed up the risk scale by nothing more than the passage of time and the arithmetic of compounding, which is exactly the opposite of the direction most people’s horizon is moving.

The check that catches this takes five minutes once a year. Write each sleeve’s current value in a column, divide each by the total, and compare against the targets you chose. If you cannot recite your targets, that is the finding.

Rebalancing bands: absolute versus relative

A band is a rule that says how far a sleeve may drift before you act, and it exists so that the decision to trade is made by arithmetic rather than by mood.

An absolute band is a fixed number of percentage points. A 5 point band on a 60 percent stock target triggers at 55 or 65 percent. A relative band is a percentage of the target weight itself. A 25 percent relative band on a 60 percent target is 15 points, triggering at 45 or 75 percent.

The two give different answers, and the illustrative example shows it. After the strong year, stocks sit at 65.5 percent, which is 5.5 points from target. That breaches a 5 point absolute band and requires a trade. It is nowhere near a 25 percent relative band, which would allow drift to 75 percent, so under that rule you would do nothing at all.

Which is better depends on the sleeve. Absolute bands are blunt on small sleeves: a 5 point absolute band on a 5 percent cash target can never trigger on the downside, because cash cannot fall below zero. Relative bands scale sensibly across sleeves of different sizes but permit large absolute drifts on the big ones. A common compromise is a relative band with an absolute cap, and the honest answer is that any written rule you follow beats an unwritten one you improvise. Our step-by-step piece on how to rebalance your portfolio covers the execution.

The tax cost of rebalancing in a taxable account

Inside a sheltered account, rebalancing is close to free: you sell one holding, buy another, and the transaction generally has no immediate tax consequence. In a taxable account the same trade generally realises a gain, and a realised gain is generally taxable in the year of the sale.

Price it on the illustrative figures. Restoring the target mix on the drifted $457,800 portfolio means selling $25,320 of stocks, because the 60 percent target of $457,800 is $274,680 and the sleeve holds $300,000. The proceeds buy $23,030 of bonds and add $2,290 to cash, which is where the whole $25,320 goes.

Suppose 40 percent of that $25,320 is gain rather than original cost, so $10,128, and suppose an illustrative 15 percent long-term rate applies. The tax bill is about $1,519. Against a $457,800 portfolio that is roughly 0.33 percent, paid in cash, to move the mix back to where you chose it.

That cost is not an argument against rebalancing. It is an argument for rebalancing thoughtfully: doing the trades inside sheltered accounts where possible, using new contributions and dividends first, and not triggering a bill for a drift of half a point. It also interacts with other tax mechanics, so our notes on tax-loss harvesting and what a wash sale is are worth reading before you place a rebalancing trade in a taxable account. Rules and rates differ by jurisdiction and change, so confirm the current position with a qualified tax professional.

Rebalancing with new money instead of trades

The cheapest rebalancing trade is the one you never place. If money is still going into the portfolio, direction of new contributions is a lever that corrects drift without selling anything and therefore without realising a gain.

The mechanic is simple. Each time you contribute, compare current sleeve percentages against targets and send the new money to whichever sleeve is furthest below its target. On the drifted illustrative portfolio, bonds sit at 30.0 percent against a 35 percent target, so contributions go to bonds until the gap closes.

Size determines how much this can do. Contributions of $18,000 a year against a $457,800 portfolio are about 3.9 percent of the balance, so redirecting all of them shifts the mix by a few points a year. That is enough to handle ordinary drift and not enough to correct the aftermath of a very strong year, which is why the technique supplements a band rule rather than replacing it.

Dividends and interest work the same way. Turning off automatic reinvestment in the overweight sleeve and directing that cash to the underweight one is a quiet, continuous rebalancing that costs nothing extra, though it does mean the income is no longer compounding inside the holding that produced it. Our note on dollar-cost averaging covers the contribution habit itself.

What allocation cannot do for you

Allocation is powerful and it is not magic, and being clear about the limits prevents a great deal of disappointment.

It cannot prevent losses. A diversified mix falls in bad years; it falls less than a concentrated one, which is the entire claim. Anyone describing an allocation as protection has overstated it.

It cannot make asset classes independent. Correlations are not fixed properties, and in severe market stress things that usually move separately often move together, precisely when separation would have been most valuable. An allocation should be built to be tolerable if that happens rather than on the assumption that it will not.

It cannot fix a savings rate. On a $400,000 balance, the difference between an assumed 5.75 percent and 6.4 percent return is about $2,600 in the first year. An extra $500 a month of contributions is $6,000. For most people in the accumulation phase, contributions do more work than the mix does, and reversing that priority is a common and expensive error.

And it cannot substitute for a plan. The allocation serves a goal, and if the goal is unspecified the mix cannot be right or wrong, only arbitrary. The calculator on our homepage exists to put a number on the goal first, and our note on the 4 percent rule covers what that number is meant to support.

Common mistakes when setting an allocation

The first is choosing funds before choosing the split, which reliably produces a collection rather than a portfolio, often with three holdings making the same bet.

The second is measuring the allocation one account at a time. Only the combined figure describes your exposure, and account-level allocations can be individually sensible and jointly wrong.

The third is treating a target-date fund or managed portfolio as a sleeve rather than as a whole allocation. Such a fund already contains its own stock and bond split, so placing it alongside separate holdings makes the true mix considerably harder to read.

The fourth is changing the allocation in response to markets. An allocation revised after every bad quarter is not an allocation; it is a series of timing decisions, and it produces the sell-low, buy-high pattern that allocation was supposed to prevent.

The fifth is ignoring what the stock sleeve actually holds. Two portfolios both described as 60 percent stocks can behave very differently if one holds broad global exposure and the other holds a handful of concentrated positions. The label is not the content.

The sixth is forgetting that human capital counts. Someone whose income depends on a single industry already carries a large concentrated exposure to it, and adding an overweight to the same industry inside the portfolio doubles a bet that is already there.

A short annual routine for your allocation

The whole of this subject can be maintained in about half an hour a year, on a date you pick in advance and keep.

Start by writing down the current value of each sleeve across every account, then divide each by the total to get today’s percentages. That single table is the entire diagnostic, and it takes ten minutes.

Compare each percentage against its target and against your band. If nothing has breached, you are finished, and the correct action is genuinely to do nothing. If something has breached, note the amount and where the correcting trade can be placed most cheaply, preferring sheltered accounts and pending contributions over taxable sales.

Then ask the one question that is not arithmetic: has anything changed about the horizon, the income, the obligations or the goal? A new dependant, a job change, a house purchase moving closer, a retirement date shifting: these change the ability side of the risk equation and are the legitimate reasons to revise a target. Market performance is not on that list.

Finally, write the date, the percentages, any action taken and the reason next to last year’s entry. That running record is what turns an allocation from a vague intention into something you can actually be held to, including by yourself in a year when it feels wrong.

The bottom line

Asset allocation is the split of a whole portfolio across broad asset classes, and it is the decision that explains most of how the portfolio behaves. The illustrative mix used throughout this explainer, 60 percent stocks, 35 percent bonds and 5 percent cash on a $400,000 balance, means $240,000, $140,000 and $20,000, carries an assumed long-run return of about 5.75 percent, and carries an assumed bad-year decline of about 26.8 percent, which is $107,200 in cash terms. Every one of those figures is invented round arithmetic, chosen to be internally consistent and easy to follow.

The three inputs that should set the split are horizon, ability to take risk, and willingness to take it, and the binding one is whichever is lowest. Age rules are starting points that disagree with each other by twenty percentage points on the same person, which is the clearest possible evidence that they are shortcuts rather than answers. Glide paths turn the intention to reduce risk into a schedule that executes itself, and asset location decides which account each sleeve sits in without changing the allocation at all.

Left unattended, the mix drifts one way. The illustrative 60/35/5 becomes about 65.5 percent stocks after a single 25 percent stock year, roughly 67 percent after ten years of the assumed returns, and about 73 percent after twenty. Bands decide when that drift is worth correcting, new contributions correct it most cheaply, and in a taxable account the correction has a price: about $1,519 on the $25,320 sale in the worked example. None of this requires a forecast. It requires a written target, an annual look, and the discipline to act on the table rather than on the news.

Dividora publishes arithmetic a reader can check rather than conclusions a reader must accept, and this explainer is offered in that spirit: educational general information only, not investment, tax or legal advice, and not a recommendation of any particular allocation, asset class, account type or product. The portfolio values, sleeve weights, assumed returns, assumed declines, single-year results, gain fractions and tax figures above are invented round numbers selected for internal consistency; they describe no real market, forecast no outcome, and no allocation described here has been shown to produce any result. Asset classes can fall together, bonds can lose value, cash loses purchasing power to inflation, and a mix that suited one decade may suit the next poorly. Tax treatment of investment income, realised gains and each account type differs by jurisdiction and changes without notice, so verify anything time-sensitive against a primary source. Before letting any of this shape a real portfolio, take your own goals, horizon, obligations and account types to a qualified financial or tax professional.

Edited by Hamza Hai, MBA · Editor

Hamza Hai is the editor of Dividora. She holds an MBA and reviews the site's articles against our editorial standards, checking that every figure is labelled for what it is, that nothing is presented as verified fact without a source the reader can check, and that the writing stays useful to a non-specialist.

How we research, write and review · LinkedIn