Is Your House Part of Your Asset Allocation? I Call Myself a 100% Stock Investor and Own Two Houses

The two largest numbers I have ever signed my name to are both buildings. Neither of them has ever appeared in a sentence I wrote about asset allocation.

So when somebody asks me is your house part of your asset allocation, the honest first answer is a confession. I have said on this blog, in public, that I hold 100% stocks and have never owned a single bond. Still true. It is also only true about the smaller half of what my wife and I own. The rest is bricks, in one country, and we are still paying off the second building, with roughly twelve years to go on that one. I have a detailed, slightly nerdy opinion about the minority of our balance sheet and no opinion whatsoever about the majority.

Earlier this year I went digging inside my index fund looking for hidden concentration and wrote up what I found. It did not occur to me to look outside it.

Here is where I land, and it is uncomfortable in both directions. A house is not a bond, and it is not pure consumption either. It is a leveraged, undiversified, single-city position you already own and cannot rebalance, and once it sits on the same page as the portfolio, most homeowners find their real mix looks nothing like the one they argue about online. Two things follow, and neither is what people want to hear: the liquid part of your money should usually stay aggressive rather than get safer, and the roof over your head should never count toward the number you plan to live on.

Is your house part of your asset allocation? Put both on one page and find out

The exercise takes ten minutes and it is not a trade. One line-up: the gross value of each property, the loan against each one, the portfolio, the cash, everything as a percentage of the total.

Gross value and loan as two separate lines, never just the equity. The best sentence I found on this whole question came from a commenter under a blog post from 2009: if someone owns a $500,000 house, he has exactly the same real estate exposure whether he owns it outright or has a $490,000 mortgage. Your equity is your slice. The building is your exposure.

Then the aggregates, because the cross-Atlantic picture is stranger than I expected.

In the euro area, the ECB's household survey (Wave 2023, released June 2026, roughly 90,000 households) puts real assets at 79.3% of everything households own: the main residence alone is 47.0% of total assets, other property another 19.1%. Two-thirds of the pile is buildings. Publicly traded shares come to 2.0%. Mutual funds, 2.8%. Bonds, 0.6%. Only 11.4% of euro-area households own listed shares at all, while 99.1% hold deposits.

Now the United States. The Federal Reserve's Z.1 accounts through 30 June 2026 show household total assets of $217.8 trillion, of which owner-occupied real estate is $49.8 trillion (22.9%) and corporate equities are $74.0 trillion (34.0%). Side by side, Americans look like stock investors who happen to own homes and Europeans look like property investors with a brokerage app.

Data: ECB Household Finance and Consumption Survey, Wave 2023 (June 2026), and Federal Reserve Z.1, Table 1, 2026 Q2. Two different measurements, shown separately on purpose: the euro-area panel is a household survey of roughly 90,000 households, while the US panel is a national-accounts aggregate for the whole household sector that is pulled upward by the wealthiest households. The median US family's balance sheet looks far more like the euro-area panel.

Except that comparison is a trap. The US aggregate is a sector total, and that sector is dominated by a small number of very rich people who own most of the equities. Drop to the typical American family and the picture flips. In the 2022 Survey of Consumer Finances the median homeowner's net housing value was $201,000, the largest in the survey's history, while the median value of stock held by families who own any stock was $52,000. Two different conditional medians on two different groups, and I want to be straight about that. The shape is still not subtle: the median US household's balance sheet looks thoroughly European.

So the argument travels. The mix people actually debate, stocks versus bonds inside a brokerage account, is a rounding error on most balance sheets on either side of the ocean.

Counting the house is an accounting question, and I wrote that how-to in how to calculate your net worth. Allocating around it is the question nobody answers.

Most of the advice here is a bookkeeping rule wearing a suit

I read every page that ranks for this question. They all land in the same place: yes, home equity belongs in net worth; no, it does not belong in your investment allocation; keep two numbers. That is a filing instruction, not an answer. Not one asks the obvious follow-up, which is so what percentage of everything you own is property, and what changes because of it?

The two popular positions are each half right, which is why the argument never ends.

"Your house is basically a bond." The steelman is decent. It pays you rent in kind every month, that stream roughly tracks inflation, the price is sticky, and it will not go to zero. The rent-in-kind part is completely real and I concede it. Then it falls apart in four places. A bond has a maturity date and a known redemption value; a house has neither. A bond has a quoted price; your house's price is your opinion until somebody bids on it. A bond does not have a roof, a boiler or a neighbour. And a bond is not usually bought with borrowed money: even at a record level of home equity, US households carry about 39 cents of mortgage alongside every dollar of equity ($49.8 trillion of housing against $35.8 trillion of owners' equity). Nobody would hold one corporate bond at 47% of their assets and call it ballast.

"A house is consumption, not an asset." The strongest version is the one the Bogleheads forum has been repeating for fifteen years, and it is cleaner than mine: an allocation is a plan for capital you can deploy, and you cannot deploy your kitchen. If you cannot rebalance it, it is not part of your portfolio.

For the primary residence I think that is mostly right. But it protects exactly one of my two buildings. The same position explicitly carves out investment property, because that can be sold if the numbers stop working. I own one of those, and I know what it costs me, because I added up the hours and it turned out to be a part-time job. It also says nothing at all about the mortgage, which is real risk on the liability side whether or not you flatter the asset by calling it an investment.

You cannot rebalance a house, and that is the actual problem

Every rebalancing rule assumes you can trade both sides. I use threshold bands and have written out the routine, and the whole mechanism quietly assumes that when something drifts 5 points out of line I can sell a slice of it.

With property there is no bit of it: there is 100% and there is 0%.

The toll is measurable. Global Property Guide puts round-trip residential transaction costs, buy plus sell as a share of property value, at 2.52% to 4.20% in Estonia, 3.65% to 10.10% in the United States, 8.50% to 15.00% in Germany, and 10.50% to 20.00% in Spain. The world ETF I buy every month has an ongoing charge of 0.14% a year. One round trip on a house costs more than several decades of holding the fund.

Data: Global Property Guide round-trip cost tables (September 2025) and the Vanguard FTSE All-World UCITS ETF factsheet (August 2026). These are not like-for-like: the property figure is a one-off toll paid when you buy and then sell, the fund figure is an annual holding cost. Spread over typical holding periods, Jordà and his co-authors put the annualised drag on housing at roughly 77 basis points against about 100 for equities, so the argument is about trade granularity rather than total cost.

Now the rebuttal, because it is a good one. Jordà and his co-authors work out that spread across typical holding periods, roughly 7.7% every ten years or so, the annualised drag on housing comes to about 77 basis points, against roughly 100 for equities at historical turnover rates. On that measure housing is no more expensive to own than stocks.

So let me narrow the claim. Owning a house is not unusually expensive. It is unusually rigid. The cost is not the toll, it is that the only trade sizes on offer are zero and everything, and the sale you will actually be forced into is the unplanned one.

The first uncomfortable answer: this usually argues for more equities, not fewer

If you have just discovered that two-thirds of your balance sheet is bricks, the instinct is to reach for safety in the part you can still touch. Sell some stocks. Add bonds. Calm down.

I think that is backwards, and I hold it with less confidence than most things I write here.

Adding bonds reduces risk in the sleeve you can steer. It does not reduce your single-city concentration by one cent. You stack conservatism on top of the conservatism you already have, in the one place you could still act. I argued the case for a 100% stock portfolio on its own merits a few months back, and here is what I left out of that post: the liquid sleeve was never the whole allocation, and once you see the whole thing, 100% stocks looks less like aggression and more like the only steering wheel left in the car.

Now the part I would be dishonest to skip: the canonical academic treatment of this exact question says the opposite of what I just said.

Flavin and Yamashita modelled the household portfolio with owner-occupied housing in it, twice in the American Economic Review, and concluded that the share held in stocks should be decreasing in the ratio of house value to net wealth. At a relative risk aversion of 3, their optimal stock-to-net-worth ratio is 0.09 for households aged 18 to 30 and 0.60 for those over 70, and they confirmed the pattern empirically in the Survey of Consumer Finances. By that model, a young mortgaged homeowner running 100% equities is badly over-risked, and I am that guy.

My reply is not a refutation. Their risk budget is a single number and my concentration is a place, and no quantity of bonds moves my exposure out of one city. But I am arguing against that paper, not around it, and you should know that before you copy me.

There is one condition where I flip completely, and it is not a percentage. If you could not service your mortgage through a long stretch without your main income, stop reading about allocation. That is the thing to fix, with cash and a smaller loan, not with a different fund.

Is your house part of your asset allocation? Yes, and it still does not count toward your number

Your total wealth and the number you plan to live on are two different quantities, and the house belongs in the first one only.

Take the euro-area median main residence, €200,000, and imagine it appreciating 30%. Net worth jumps €60,000. Monthly freedom budget changes by exactly zero, unless you are genuinely prepared to leave. And people are not. In the second quarter of 2026, American homeowners with a mortgage were sitting on $17.9 trillion of equity and borrowed against less than 0.1% of the equity they could have drawn on. The wealth decumulation literature going back to Venti and Wise finds the same among retirees: housing wealth mostly does not get spent. People treat the house as unspendable because, in practice, it very nearly is.

So the rule I use: run the projection twice, with the roof and without it, and the one without it is the plan. A plan that only reaches its target by counting the house has not reached its target.

I build the tool that does this for me, so discount my enthusiasm accordingly. It is four rows in a spreadsheet otherwise, and the point is that both numbers stay visible, because the flattering one has a habit of being the only one you look at.

Your house is a bet on one street, and the data on that is brutal

The strongest argument against everything I have written is that residential property has been a perfectly good long-run asset, and it comes with numbers.

Jordà, Knoll, Kuvshinov, Schularick and Taylor priced 16 countries over 145 years. Housing returned 7.05% real a year on the arithmetic mean against equities' 6.89%, at a standard deviation of 9.98 against 21.94. Roughly the same return, half the volatility. The authors call it a puzzle, not a rounding error, and I am not going to wave it away.

Data: Jordà, Knoll, Kuvshinov, Schularick & Taylor, "The Rate of Return on Everything, 1870–2015", Table 3. These are national housing indices, not the house you own: the same authors note that US ZIP-level housing volatility is about twice the aggregate, which would roughly equalise the risk-adjusted returns of equities and housing for someone holding a single undiversified house. Measured at the level of actual property accounts and net of costs, Chambers, Spaenjers & Steiner (Review of Financial Studies, 2021) find long-run real residential returns of around 2.3%.

The hinge is a sentence in their own paper. US housing volatility at the ZIP-code level runs about twice the aggregate, which they say would "about equalize risk-adjusted returns to equity and housing" for an investor holding one undiversified house. Nobody owns national housing. You own one building, on one street. And when Chambers, Spaenjers and Steiner read 82 years of actual property-level accounts from four Oxbridge college portfolios, residential came out at roughly 2.3% annualised real total return net of costs, against Jordà's 6.61% geometric figure from national indices. The gap is running costs, which the indices never subtract and the colleges' bookkeepers did.

Then the chart that settled it for me. Eurostat's inflation-adjusted house price index, same currency, same methodology, same sixteen years from Q1 2010 to Q1 2026: Portugal up 94.1%, Austria up 46.9%, Germany up 29.5%, euro area up 13.4%, France up 1.0%, Finland down 23.2%, Italy down 23.8%. A 118 percentage point spread inside one currency union. The average describes almost nobody.

Data: Eurostat deflated house price index (tipsho30), index base 2015 = 100, deflated using consumer price inflation. All countries shown are euro-area members, so there are no currency effects.

The American dispersion is just as loud. Cotality reports that since 2020 the average Connecticut homeowner gained about $239,000 in equity and the average Texan gained $58,000. Neither of them made a decision.

That is the risk. Not that houses are bad. That the one you own is a single draw from a distribution that wide.

What I changed, and the three things I did not

Small, boring, and mostly a measurement change.

One page instead of two. Both properties at gross value, each loan as its own line, the portfolio and the cash beside them, percentages down the right. Reviewed once a year, not monthly, because there is no price feed on a house and checking a number you invented is anxiety with extra steps.

A deliberately conservative value for both buildings, from a rule I wrote down before I knew what it would produce: the last comparable sale I can actually point to, not the cheerful top of an online estimator's range, and never revised upward more than once a year. That rule exists because of an embarrassing minute at our last annual review. My wife asked what the second house is worth now. I answered instantly, with a confident number, then could not tell her where it came from. I had picked a figure I liked the sound of, some time in the previous two years, and the fund on the line below it was being priced by an actual market while I did it.

The three things I did not do, because this is a measurement change and not a trade recommendation: I did not sell anything, did not touch the standing order, and did not add a single bond.

We also did not undo the real mistake, because you cannot. Buying the second house was the largest allocation decision my wife and I have ever made, and we made it in an afternoon without once using the word allocation. I had my clever phase, rode one lucky pick into a crypto cycle that sat me down and explained I am not a genius, then went out and put a large, leveraged, single-city bet on the balance sheet while congratulating myself for buying a cheap index fund with the leftovers. If you are about to buy, you are making the biggest allocation decision of your life in one transaction, at a cost of 3% to 20% of the value to reverse. It deserves at least the evening you would give to choosing a fund.

The objection I cannot fully answer

If you will never sell and never move, all of this is academic.

That is the strongest thing anyone can say to me, and it deserves its full weight rather than a strawman. Price risk only bites if you have to transact. If you intend to die in that house, the only variable that matters is whether you can pay the mortgage through a bad decade, and if local prices fall the replacement house falls too, so staying put is close to self-hedging. The blogger whose page on this question has outranked everyone for seventeen years puts it plainly: if he never moves, house prices do not much matter to him.

Most of that is right. Genuinely.

What is left is smaller than my argument would like, but it does not go away. "I will never sell" is a forecast about your own life, and life supplies job offers, separations, births and ageing parents. The median American seller in NAR's 2025 data had owned for eleven years, an all-time high, and eleven years is not forever. It says nothing about a second property, which even the opposing camp counts as a portfolio position. And leverage is still leverage, which bites hardest on the day you least want to transact.

There is also a turn I did not expect when I started writing this. If the house really is permanently untradeable, that strengthens the case for keeping the liquid sleeve aggressive rather than weakening it. When most of what you own cannot move, the small part that can is the only allocation decision you still get to make.

So I have not changed a position. I changed what I look at once a year, and the output is a number I now know rather than an order I placed. If you want my short answer to is your house part of your asset allocation, it is yes for the measurement and no for the plan, and the gap between those two is where most people quietly overstate how free they are.

Put the buildings and the fund on one page and read the percentage off the right-hand side. It may tell you nothing you did not already suspect. Mine did not, and I still think about it every month when the standing order goes out.