***The report below is a combination of excerpts from recent reports out of our Financials vertical, let by Josh Steiner. If you are an institutional client or prospective client and aren’t yet receiving Josh’s work on housing, credit, and the financials, please email to learn more about how we can get you setup.***
Case-Shiller Falls Month-Over-Month and Decelerates Rapidly Year-Over-Year
The Case-Shiller Home Price Index fell 0.21% month-over-month in August (non-seasonally adjusted), versus up 0.65% in July. On a year-over-year basis, growth decelerated to 1.6% in August versus 3.1% in July.
Next Month Will Get Worse, Potentially Much Worse
It's critical to understand the timing associated with the Case-Shiller series. The printed number is a 3-month rolling average released on a two-month delay, so the August release today is the average of June, July, and August. Case-Shiller measures closing activity, which tends to lag signing activity by 1-2 months. To compare Case-Shiller to the MBA Mortgage Purchase Applications Index, we should look at applications from April/May/June. Thus, today's Case-Shiller print is still capturing a benefit from the strong April purchase activity driven by the tax credit. Next month, the September print will reflect signing activity from May/June/July, which will be substantially weaker than today's release. Specifically, the average of May/June/July is 13% lower than April/May/June from a demand perspective. With supply remaining at historical highs, we expect prices to increasingly come under pressure. The chart below demonstrates.
The following chart shows Case-Shiller home price data on a month-over-month basis. As we've highlighted previously, by S&P's own admission, investors should not rely on the seasonally adjusted (SA) data as their seasonal adjustment factors are essentially unreliable. Rather, investors should rely on the non-seasonally-adjusted data as a better indicator of underlying trends. It's worth emphasizing that the Case-Shiller series does have a notable seasonality - specifically, it generally improves sequentially through April, May, and June - so the NSA data has its own shortcomings.
The chart of year-over-year price change below shows another deceleration in August.
Existing Home Sales Rise as Prices Fall
Existing Home Sales rose 9.7% to 4.53 million (seasonally adjusted annualized rate) in September. As Existing Home Sales are a lagging data series, it is still benefiting from a rebound off of the post-tax-credit lows. Below we show charts of existing home sales and median prices.
Inventory Remains Near Record Highs
Inventory rose sequentially to 4.04 million, though was reported to bedown 1.9% after an upwardly-revised August print of 4.12 million homes (revised from 3.98mn units). A higher sales rate brought months supply in from 11.6 to 10.5 months. While this is the second sequential improvement in the months supply series, by historical standards, 10.5 months is consistent with the 2008 highs. Meanwhile, the median price of Existing Homes fell 3.9% month-over-month from $178,600 to $171,700.
New Home Sales Rise 6.6% to 307k - Should We Get Excited?
Don’t be fooled into thinking 307k is a strong number. In fact it is right in line with the average we’ve seen post the tax credit expiration as the charts below show, and it remains consistent with our cumulative displacement theory (republished below) that new home sales will need to remain around the 300k level for the better part of the next ten years, much to the disappointment of those who’ve bet on a snappy recovery over the next 12-24 months.
New Home Sales rose 6.6% to 307k SAAR. Is this a cause for celebration? Remember that it was just six months ago that a then-record-low print of 300k caused significant angst and a material selloff. Since then the numbers have remained in this 300k range. Expectations appear to have come a long way. To summarize our cumulative displacement theory, there was an epidemic of overbuilding during the bubble, which will take a very long time to work off. Using a sales rate of 300k, we calculate that sales would have to continue at this level for ten years for the cumulative displacement from the mean to return to zero. Yes, new home inventory is very low, but we don't see sales rebounding anytime soon.
The hangover from the tax credit expiration appears to remain more serious than the pull forward effects. As the following chart shows we are now into our fifth month of unprecedented weakness in new home sales, which begs the question: is there something else going on, something structural? We think the answer to this question is yes, and we explain why we think that in our cumulative displacement analysis below.
Taking a longer-term view back to 1993, new home sales continue to decline. The yellow line in the following chart shows the rolling six-month average - the housing equivalent of the 200DMA. This trend line shows no sign of improvement - indeed, it is worsening.
Joshua Steiner, CFA