Suneet Singal on what an affordability index does not say

Suneet Singal’s housing entries keep returning to the same warning, and an affordability index is where the warning is easiest to ignore, because the index looks like an answer. It compresses a price, an income, and a set of financing assumptions into one number so a market can be compared with itself over time or with another market. Under the publisher’s own definition, a value of 100 is usually the point at which a stated typical income just covers the payment the index assumes on a stated typical home. That is a real piece of information about those assumptions. It is not a statement about your income, your debts, your down payment, the house you can actually find, or the tax bill on a California parcel. This entry explains the compression and the omissions. It does not quote this month’s index, and it is not personal financial advice.

What the index is trying to compress

An affordability index answers a constructed question. Take a typical home price, often a median. Take a typical income, often a median family or household income. Take a mortgage rate and a set of assumptions about the down payment and about how much of income may go to the payment. Compute the income required to qualify under those assumptions. Compare the typical income with the required income. Scale the comparison so that 100 means the two incomes match, if that is how the publisher defines the scale. The National Association of Realtors publishes research of this kind, and the methodology note on its research pages is the definition that counts. This paragraph is a map of the idea. It is not a substitute for the note, and a publisher can revise the note.

The compression is the reason the index exists. A single number can be charted. It can show that a market became harder or easier to enter under a stable set of assumptions. It lets a reader talk about price, rate, and income without listing all three in every sentence. The cost of that convenience is that the three, plus the assumptions, disappear into the result. A reader who quotes the result and cannot say which assumption did the work does not yet understand the result. Suneet Singal’s entry is about recovering the pieces before the number is repeated.

Different publishers build different indexes. A housing-cost burden from the Census Bureau, such as a share of households paying more than a stated fraction of income for housing, is not the same object as a trade association’s affordability index. One is a distribution of actual households in a survey. The other is a comparison of medians under mortgage assumptions. Both can be called affordability measures in ordinary speech. They will not move together in every period, and they should not share a sentence unless the sentence says which is which.

The moving parts

Hold the methodology fixed and the direction of each part is straightforward. A higher median price, with income and the rate unchanged, raises the income required to qualify and lowers the index. A higher mortgage rate does the same, because the payment formula in the interest-rate entry turns a higher rate into a higher principal-and-interest payment on the same loan. A higher median income raises the index, because the numerator of the comparison grew. When all three move in the same release, the index can rise because income rose even while prices rose, or fall because rates rose even while incomes rose. The release, not the headline verb, tells you which story you are in.

The assumptions that are not in the headline matter as much as the three moving parts. What down payment is assumed? If the index assumes a large down payment, it describes a buyer who has that cash, not a buyer who does not. What share of income is treated as the maximum payment? A qualifying ratio is a modeling choice. It is not a budget you have agreed to, and it may or may not include taxes, insurance, and mortgage insurance. You have to read the note to see what the payment contains. What rate is used, and is it the same object as FRED series MORTGAGE30US? Sometimes a methodology uses a different rate series. Saying “rates” without naming the series imports the wrong history.

This page does not lock those assumptions to a remembered percentage. Methodologies are revised, and a remembered 20 percent or a remembered income share might match the current note or might describe an older edition. Open the note. Copy the assumptions into your own words beside the index value. If you cannot find the note, do not cite the value. An index without its assumptions is a number that can be graphed and cannot be explained, which is the condition this entry is written to prevent.

What it does not say about a household

The index does not say what you earn. A median income is the middle of a distribution. Half of the measured households are on either side of it. A household far below the median can be in a market whose index is above 100 and still be unable to qualify for the median-priced home. A household far above the median can be in a market whose index is below 100 and still have room. The index was not built to locate you in the distribution. It was built to compare the middle with a payment.

It does not say what you owe. Student loans, car payments, and other debts change a lender’s view of capacity and do not appear in a simple median-income comparison. It does not say what you have saved. The down-payment assumption inside the index is not your savings account. The down-payment entry explains how cash, loan size, and mortgage insurance interact in a real file. An index that assumes the insurance away, by assuming a large down payment, will look more affordable than the file of a buyer who will be charged the insurance. Those are different products. The index should not be asked to price both.

It does not say whether you will be approved. Credit history, employment history, the property type, and the lender’s overlays are underwriting facts. The Consumer Financial Protection Bureau’s owning-a-home material describes the documents in that process. None of them is an affordability index. A reader who treats a published index as a preapproval has confused a research product with a loan decision. The research product can still be worth reading. It is worth reading as research.

The median house is not a house you can tour

Even for a household that earns the median and has the assumed down payment, the index does not say that a suitable house is for sale. The median price is a statistic of sales or of values in a defined universe. The house at that price may be in a different neighborhood from the job, in a different condition from the one the household can accept, or simply not listed this month. Inventory, days on market, and the mix of what sold are other tables. The National Association of Realtors research page is a portal to more than the affordability series. A sales count and an affordability index can tell opposite emotional stories about the same month, because one is about how many homes changed hands and the other is about a ratio of income to a payment.

Geography hides inside the median. A national index can look calm while a metropolitan index moves sharply, and a metropolitan median can hide a county or a city that a household actually lives in. The Sacramento entry is the local demonstration: city QuickFacts, a county boundary, and a metro index are different objects. An affordability measure for a named metro is not a measure for every city inside it. Quote the geography in the table title. If the table says the United States, do not say your city “is affordable” because the national index is above 100.

Condition and insurance do not hide inside the median either, except to the extent that they influenced the prices that were paid. A house with a failing roof can sell below the median and still be a bad bargain once the repair is priced. The inspection entry is the companion. An index will not walk the roof. Using the index as a substitute for the inspection, the appraisal, or the insurance quote is how a market statistic gets promoted into a due-diligence report. It is not one.

How to read a move in the index

When the index changes, ask which input moved. A release that decomposes the change into price, rate, and income has given you the sentence. A release that publishes only the index has given you a reason to open the underlying price series, the rate series, and the income series. The FHFA house-price index is one price series, with the index caveats in the data-reading entry. FRED’s MORTGAGE30US is one rate series, and only if the methodology uses a 30-year fixed conventional average. Income may come from a Census series or from another source the note names. Match the source to the note. Do not explain an NAR index with a random chart that uses a different price concept and then act surprised when the stories diverge.

A one-month move can be noise or a turn. The data-reading entry’s advice on revisions and on seasonal adjustment applies here. Affordability can be published as a seasonally adjusted series or not, and monthly median prices move with the mix of homes that sold. A month in which the homes that sold were larger can look like a price spike and then like an affordability collapse, without any change in the price of a constant-quality house. The FHFA repeat-sales index is one attempt to hold quality more constant. It will not match a median sale price month for month. If you cite both, say both.

Do not convert a change in the index into a change in “how many people can buy.” That conversion requires a distribution of incomes and a distribution of prices, not a ratio of two medians. The index can fall while the number of sales rises, if more people buy cheaper homes or if more buyers are above the median. Sales are a count. The index is a ratio. They are neighbors in the housing conversation and they are not substitutes.

Housing cost is more than the payment inside the index

Whatever payment the methodology uses, a household’s housing cost also includes the items the methodology left out. Property tax in California depends on assessed value under Proposition 13, not on a generic fraction of the current median price. The property-tax entry tells you to read the Board of Equalization before you invent that bill. Hazard insurance has been a large and uneven cost in recent years, and a national index built on principal and interest will not show a local insurance quote. Mortgage insurance depends on the loan-to-value in the file. Utilities, association dues, and maintenance sit outside most affordability indexes entirely. The rent-or-buy framework is where those rows belong. The index is not that framework. It is one input a person might misuse as if it were the framework.

Closing costs and the cash to close are outside the index for a different reason. They are stock, not flow. A household can “afford” the monthly payment under a qualifying ratio and still not have the cash to close. The closing-cost entry is the reminder. An index that assumes the down payment also assumes the household somehow assembled it. The index does not say how.

Rent belongs in the conversation and usually belongs in a different measure. A homeownership affordability index does not say whether renting the same shelter is cheaper this year. Some publications do publish a rent comparison. If they do, read the rent definition. A median gross rent from the Census and an asking rent on a portal are not the same observation. The data-reading entry’s universe rule applies. Do not let the word affordable glue a rent statistic and a purchase index together without a sentence that shows the seam.

How to use the index without making it confess

Use the index as a headline that you immediately unpack. Write down the value, the date, the geography, and the three inputs as the release states them. Write down the down-payment assumption and what the qualifying payment includes, copied from the methodology note. Then write one sentence the index does not authorize. The sentence will be something like “this does not say whether a named household qualifies, and it does not include a California tax bill or an insurance quote.” If you cannot write that sentence, you are about to overclaim.

Place the index next to the series it is made of, not next to a personal anecdote that it cannot see. Place the personal facts, when you have them, next to a Loan Estimate and a tax bill. The two stacks can inform one decision. They stay in two stacks so a reader can tell a national or metro construction from a file. Suneet Singal publishes the stacks separately on purpose. An affordability index is a useful compression of a market under stated rules. What it does not say is most of what a household still has to find out. The methodology note tells you the rules. The rest of this housing reference tells you where the missing rows are kept.

What this entry does not claim

Suneet Singal’s published professional background is in real estate and corporate finance. Company filings linked from this site record a chief executive officer departure from FC Global Realty Incorporated effective 2 January 2018 and an acting chief executive officer appointment at First Capital Investment Corporation effective 12 September 2017. The Suneet Singal Scholarship, Grant and Giving websites describe him as the founder of those programs. This entry explains a general subject. It is not a private track record, a current job title, an investment result, or a quote.

Continue with how to read a housing data release, a rent-or-buy framework, and California property tax basics.