The MBA’s weekly mortgage application survey is one type of data that subtly reveals the true state of the economy but doesn’t make the nightly news. The mortgage bankers association publishes a set of statistics every Wednesday morning, which the housing sector examines like a physician examining a patient’s pulse. Mortgage applications increased 1.7% from the previous week, according to the most recent verified data for the week ending May 8. a tiny quantity. However, even a slight increase is significant in the context of spring 2026, when buyers are anxious and rates are tense.
The headline number might be misleading, so it’s important to be clear about what these indices really measure. Every refinance and buy loan that passes through the system is tracked by the Market Composite Index. The 1.7% increase was the result of a 4% increase in the Purchase Index, which is the component that is most important for the real housing market because purchase applications reflect individuals who are attempting to purchase homes rather than homeowners who are rearranging their debt. Applications for the week ending May 1 had decreased by 4.4%. Instead of having a clear direction, what you’re actually witnessing is choppiness—a market that fluctuates weekly. The true tale is that instability.
It’s a bit contradictory, but I think the refinance photo is the most telling. It may not seem like much, but the Refinance Index actually decreased 1% week over week. However, compared to the same week last year, it was 28% higher—a truly significant year-over-year increase. You can learn something about the trajectory of rates from that gap. Early in 2026, homeowners who had purchased during the higher-rate years of 2023 and 2024 hurried to refinance as the 30-year conforming rate fell to about 6.09%, a 16-month low. During that period, refinance activity reached its highest level since 2022, according to MBA economist Joel Kan. Because of how much worse the rates were last spring, the annual comparison is still impressive. On the other hand, the week-to-week softness indicates that rates are gradually rising again.
And as they always are in our industry, rates have been crucial. For loans of $806,500 or less, the 30-year conforming mortgage has fluctuated in the 6% range throughout 2026, falling to those February lows before rising as Treasury yields increased. The cause of the increase is already well known: the war in Iran and the ensuing shock to oil prices raised inflation expectations, which in turn caused the 10-year Treasury note to rise. Mortgage rates then followed, as they normally do. Applications plummeted, falling more than 10% in a single week during a period in March when rates reached a five-month high of 6.43%. Since then, the recovery has been genuine but precarious—exactly the kind of circumstances that result in whipsawing from 1.7% up to 4.4% down.
Reading these figures every week makes it difficult to ignore how totally dependent on one factor the American housing market has become. In the past, rates, inventories, prices, and incomes all worked together to determine affordability. Nowadays, it’s basically about rates, and rates are mostly about Treasury yields, which are primarily about geopolitics and inflation and have nothing to do with whether a young family in Ohio can afford a starter house. Depending on what happens in the Strait of Hormuz, a buyer who could easily afford the payment at 6.09% is priced out at 6.5%. We have a peculiar method of managing the housing market.
Despite the current uptick, the buying side in particular has been persistently sluggish in comparison to the past. Purchase applications are still well below the pre-pandemic average, even with the 4% increase in the last week. Prospective purchasers are in a similar situation: even if rates have dropped from their peak, the combination of high prices and mortgages with interest rates above 6% keeps the monthly payment burdensome. This is because home prices never truly decreased following the rate increase. Housing experts believe that basic math, rather than a lack of desire, is holding back buy demand. People desire to make purchases. They can’t because of the numbers.
A little disclaimer regarding the data itself is worth mentioning. The most recent totally confirmed reading is still the May 8 survey. The statistics for the week ending May 15, which were the initial premise of this essay, would normally be issued around May 20. Economists typically monitor the trend over a month or two rather than responding to any one week because these weekly figures are equally erratic and seasonally noisy, prone to abrupt swings around holidays and rate changes. While positive, a 1.7% rise is not a game-changer. The previous week’s 4.4% decline wasn’t a collapse. The pattern, not the print, is where the signal resides.
