The CBO releases forward-looking projections of NIPA profits twice a year. In the absence of some reason to do otherwise, I use the CBO’s estimates of NIPA profits in the daily version of my S&P 500 valuation model. While the CBO’s forward-looking NIPA profits estimates often prove to be quite accurate, no one is endowed with the gift of prescience and at the moment there is (in my view) a reason to at least make some adjustments the CBO’s profits projections. The CBO’s most recent estimates, which were released on February 5, 2013, had the annualized level of NIPA profits as of 4Q 2012 at $1.966 billion. While the CBO does not publish an estimate of NIPA profits as a percentage of nominal GDP per se, such can be easily calculated (by using the CBO’s nominal GDP estimates in conjunction with its NIPA profits estimates). That figure was 12.37% for 4Q 2012 per the CBO’s February 5, 2013 projections. When the Bureau of Economic Analysis’ released its first estimate of 4Q 2012 (on March 28, 2013) the actual annualized level of NIPA profits as of 4Q 2012 was reported to have been $2,013 billion, or $47 billion higher than the CBO’s estimate. Based on actual data in hand at present, NIPA profits as of 4Q 2012 equaled 12.69% of nominal GDP, or 32 basis points than the CBO’s (imputed) 4Q 2012 estimate of 12.37%.
I will spare you the tedium of a detailed explanation of the methodology I employ to come up with the NIPA adjusted profits estimates I am now feeding into my daily S&P 500 model (any client who really cares can call me and I’ll be happy to walk you through it). A short summation is that I have used consensus forward GDP estimates and have adjusted the CBO’s NIPA profits projections to account for the 4Q 2012 “miss” in the CBO’s (imputed) NIPA profits to nominal GDP ratio.
The most recent actual NIPA profits data available at this time is $2,013 billion as of 4Q 2012. My current estimates for the annualized level of NIPA profits in 2013 are as follows:
1Q 2013: $2,006.65 billion
2Q 2013: $2,019.60 billion
3Q 2013: $2,035.33 billion
4Q 2013: $2,062.99 billion
I want to emphasize three points here. One: These are just estimates that may well prove to be significantly off the mark – in any event they are quite unlikely to be exactly correct. Two: These estimates are what feed my S&P 500 valuation model, and like any model, “garbage in” (if these estimates ultimately prove to be garbage) will result in “garbage out” (the model will yield false S&P 500 valuation estimates). Three: Even though I present some precise numbers and levels of the S&P 500 (in some cases to 2 decimal places) in the analysis that appears below do not be misled – the nature of the analysis is in fact not that precise.
I offer my apologies for the all the tedious preliminaries and caveats. The point all this has been leading up to is this: Assuming my estimate of the annualized level of NIPA profits in 2Q 2013 is correct (or close to being correct) last week the S&P 500 may have reached a major “relative valuation” resistance level and for the third time since 2009 may have formed a “relative valuation” peak. Figure 2 illustrates (from January 2, 2009 through April 16, 2013) the daily closing level of the S&P 500 as a percentage of the quarterly model value for the S&P 500. Note that a relative valuation peak occurred on June 12, 2009 at 86.38% of the model value and another on February 18, 2011 at 86.58% of the model value. Based on my estimate of the annualized level of NIPA profits for 2Q 2013 ($2,019.6 billion) the model value of the S&P 500 in 2Q 2013 is 1832.65. Last Tuesday (April 11, 2013) the S&P closed at 1593.37, or at 86.94% of the model value – close enough to the two prior relative valuation peaks to merit notice, especially since the rally subsequently failed.
By definition, a “peak” means a level that is not surpassed. So if in fact a relative valuation peak was formed at 86.94% of the model value last Tuesday, then such implies that the S&P 500 will not surpass the 86.94% relative valuation mark, which at least for 2Q 2013 would translate to a closing level above 1593.37. As shown above, my quarterly estimates call for a small increases in the annualized level of NIPA profits in 3Q 2013 and 4Q 2013. An 86.94% relative valuation level would translate into a closing level of 1606.36 in 3Q 2013 and 1626.19 in 4Q 2013.
My longer-term expectation has been, and continues to be, that at some point in the future “mean reversion” will occur and the actual level of the S&P 500 will once again rise to (at least) 100% of the model value. The immediate question at hand is whether the world, at this moment, looks safe enough (and / or whether central banks are now providing sufficiently overwhelming amounts of liquidity) that the S&P 500 can do what it has not been able to do over the past four years and break above relative valuation resistance (at roughly 86% to 87% of the model value). I do presume to know the answer (though I ardently wish that I did). I am merely pointing out that based on my model and my 2Q 2013 NIPA profits estimate, the S&P 500 appears to have reached an important potential inflection point.
RISKS: I want to reemphasize the difficulty presented when employing my model in real-time due to the necessity of using estimates of NIPA profits, and that the above analysis is based on estimates. Regular readers will perhaps recall that back in early March of this year, when I was using unadjusted CBO NIPA profits estimates, it had appeared to me that the long-standing 86.58% relative resistance level had finally been breached to the upside (when the S&P 500 closed above 1530). I now have reason to think that the NIPA profits estimates I was using at the time may well have been too low; this illustrates the point that the validity of the analysis presented above ultimately hinges on the accuracy of the NIPA profits estimates I am using (and the accuracy of such estimates is uncertain).

• MV = PQ
• The Negative Output Gap is Now Our Friend
• The S&P 500 is Not Overvalued

1. 1.      MV = PQ

Money * the Velocity of Money = Price * Quantity.

Where:

P*Q = Nominal GDP, and Q = Real GDP.

V is (in my view) quite frankly a bit of a fudge factor, but assuming V is stable what this simple monetarist equation posits is that if the supply of money increases, nominal GDP (P*Q) will increase.  “Money” is a slippery concept, and thus there is always a question of just what measure one should use for M.  I am inclined to use the broadest measure I am aware of, which is the Center for Financial Stability’s Divisia M4 index, or DM4, which in addition to being a very broad measure of “money” goes beyond simple aggregation by weighting each individual component of DM4 for its current level of “money-ness”.

As is illustrated in Figure 1, as of December 2012 the year-over-year measure of DM4 was +6.80%, which was the fastest year-over-year growth rate since January 2009 (+7.96%).  As of January 2013 (the most recent reading available) the year-over-year measure was a still high (by recent standards) at +6.1%.

So far so good, but if M is fact increasing at a faster pace, it is possible that such will manifest as faster growth in P (inflation) rather than faster growth in Q (real GDP).  Enter concept #2.

1. 2.      The Negative Output Gap is Now Our Friend

While a faster pace of M growth could manifest more in P than in Q, there’s a reason to believe that in the present circumstance, P will stay tame.  As is illustrated in Figure 2, as of 4Q 2012 the real GDP output gap was still quite negative at -5.85% of the CBO’s (the Congressional Budget Office’s) estimate of potential real GDP.  As is also illustrated in Figure 2, historically a negative real GDP output gap has been associated with a slower / decelerating pace of inflation (and vice versa).  Again, using the CBO’s estimates of potential real GDP, as of 4Q 2012 the real GDP output gap was -5.85%.  At its worst in the wake of the 2001 recession the negative output gap reached -2.1% (in 1Q 2003).  I calculate that if actual real GDP grows at a +3% annualized rate in 2013 and at a +4% pace thereafter, the negative output gap would not be near -2.1% until sometime in the first half of 2015 (-2.38% in in 1Q 2015 and -1.93% in 2Q 2015) and the output gap would not turn positive until 3Q 2016.

What this means is that while the Fed is quite unlikely to be adding more stimulus by increasing the size of its balance sheet this whole time, if real GDP can in fact grow at a +3% to +4% pace the Fed is likely to just “let it rip” for a couple of years before starting to actually tighten.  A faster pace of economic growth, if such can be achieved, has a lot of unencumbered headroom.

1. 3.      The S&P 500 is Not Overvalued

Using the CBO’s estimate of the annualized level of NIPA profits in 1Q 2013, the model value (per my long-term NIPA profits-driven S&P 500 valuation model) of the S&P 500 in 1Q 2013 is 1766.54.  As is illustrated in Figure 3, if in fact the S&P 500 rises by a couple of points from yesterday’s closing level to reach the all-time high of 1565.15 (first reached on October 9, 2007) the actual level of the S&P 500 would be just 88.6% of the model value.  When the S&P 500 first reached 1565.15 in October 2007 its relative valuation was 116.8%.  At the March 24, 2000 peak when the S&P 500 closed at 1527.4 the relative valuation was 214.3% of the model value.  While 1565 on the S&P 500 may look high relative to the recent past, 1565 is not high relative to the current level of corporate profits.

Risks:  Obviously the risks to any bullish thesis are myriad, including geopolitical risks, a major mistake by policymakers, fiscal austerity starting to bite in the U.S., etc.  Beyond the multitude of such amorphous risks, here I will highlight two risks more specifically:  The first is the fact that I’ve even written this bullishly-leaning missive.  Everyone (or lots of people, anyway) thinking bullish thoughts is often a bearish sign.  The second specific risk I’ll highlight is in regards to money growth.

The inconclusive Italian elections notwithstanding, Europe has been relatively quiescent since Draghi pledged to do whatever it takes late last summer.  Nevertheless, Europe’s problem (which, in my view, is in essence the Euro itself) has not been fixed.  The biggest risk from Europe isn’t that any particular country (or even the Eurozone, itself) has negative GDP growth, nor is it even the dissolution of the single currency, per se.  The biggest risk is that events in Europe, or fear of potential events in Europe, cause the global financial to freeze again (or partially freeze), an occurrence which at best would probably result in a very slow pace of broad money growth and at its worst cause another contraction in broad money.

I’m still sifting through (and frankly trying to make sense of) all the details in the recent broad money data, but my preliminary conclusion is that the real swing factor behind the recent growth in broad money has been growth in the components generated by the “shadow” banking system (items such as repo and commercial paper).  My fear is that the net result of new financial regulations both in the U.S. and Europe could seemingly (?) potentially (?) be collateral shortages, shorter chains of re-hypothecation, etc. in the repo market and, consequently, a curtailment of the credit (and thus the “money”) created by the shadow banking system.  These are probably more intermediate to long –term, as opposed to short-term, concerns.  Nonetheless, the possibility of adverse, and at this point largely still unknown, unintended consequences from the implementation of well-meaning financial regulations remains a concern.

Finally, I’ll note that while deleveraging has seems to have progressed well enough, “money” creation is the result of credit creation.  Have balance sheets, in aggregate, really been repaired to an extent that will allow for a significantly positive pace of credit, and thus money, creation?  Since I don’t know the answer, I will assume that the proof is in the pudding. That is, as long as the broad money data show growth, I’ll assume that the necessary preconditions for that money growth to have occurred (the ability of balance sheets, in aggregate, to absorb the credit created) did (do) in fact exist.  However, if I see money growth slow again, I’ll start to be less sanguine.

Demand for mortgages can be increased by either lowering mortgage rates or by loosening underwriting standards.  Mortgage rates are to some extent hostage to Treasury rates.  That said, it’s possible that as the Fed continues to hold a larger and larger percentage of the total stock of agency MBS, spreads over Treasuries will compress, resulting in either outright lower mortgage rates, or at least mortgage rates that are lower than would otherwise be the case.  Loosening underwriting standards (something originators arguably have more control over) increases aggregate demand for mortgages by expanding the pool of potential homebuyers (or, as has been seen with the government’s HARP program, expanding the pool of potential home refinancers).

The MBA’s composite index (shown in Figure 1) is composed of the purchase index and the refinancing index.  As is illustrated in Figure 2, despite the recent downturn in the 13-week moving average of the purchase index, the overall trend in the purchase index is still appears to be up.  The trend in the refi index, on the other hand, now appears to be down despite the boost it has (at least heretofore) been getting from the loosening in HARP refinancing standards (HARP 2.0) last year.  As is illustrated in Figure 3, refi as a percentage of total mortgage applications has declined from a recent peak of 84.0% in the week of December 7, 2012 to 75.7% last week.  There are a lot of borrowers who could qualify for HARP that have not yet taken advantage of the program, so it’s possible that refinancing activity will pick up if more people start participating.  It’s also possible that Treasury rates will drop again, resulting in lower mortgage rates and an increase in refi activity (as well as purchase activity).

The above having been said, even if Treasury (and thus mortgage rates) decline again, there comes a point when all the people who are going to refinance at any given interest rate will have done so.  It’s also worth noting that while refinancing activity changes the characteristics of the existing stock of agency MBS, it does not increase it.  The only way to increase the stock of MBS is to increase the amount of mortgages that exist via an increase in the number of mortgages originated for purchase.  If rates decline again, that would presumably result in an increase in mortgage purchase activity, but my best guess is that the stars are now aligning in such a way as to lead to a loosening in mortgage purchase underwriting requirements (a general loosening in mortgage credit).  If I’m right (and depending on the extent of the loosening in mortgage credit) such would probably not only help keep the recovery in housing market activity on track, but could potentially even help to speed the pace of the recovery.

Now, if you have listened to the speeches (and in the case of Bernanke, news conferences and Congressional testimony) of FOMC members, you know that the Committee’s forward guidance is actually more nuanced than simply “we’ll raise rates when the unemployment rate hits 6.5%”. The Fed is (rightfully) going to consider all the labor market data, including the behavior of the labor force participation rate, when contemplating the timing of the first tightening. My own sense is that the Fed more likely to stay easy in a scenario in which job growth is relatively robust, but in which the unemployment rate stays stubbornly high due to a significant rise in the labor force participation rate, than it is to tighten is a scenario in which job growth is relatively mediocre but the unemployment rate nonetheless falls to 6.5% due mainly to a continued decline in the labor force participation rate. That having been said (and nuance aside) the fact remains that the Fed has made a 6.5% unemployment rate a “policy beacon”. Thus it is probably worth contemplating the future course of the labor force participation rate, and what effect such may have on the unemployment rate (and by extension, at least the broad outline of future Fed policy).
Figure 1 presents a long-term view of the labor force participation rate. As of February 2013 the labor force participation rate was 63.5% (as noted above, 63.524% to be exact). Note that after rising rapidly starting in the mid-1960’s, the participation rate more or less flattened out in the late 1970’s and early 1980’s, a period in which 63.5% (or close to it) acted as a floor. If Figure 1 was the chart of a stock, I’d say there was technical support at or around 63.5%.
Figure 2 presents a shorter-term view. While the labor force participation rate could certainly keep declining, it looks to me as if the labor force participation rate is flattening out and (to anthropomorphize) is “trying” to bottom at 63.5%. Given that job growth (at least as recorded by the Establishment Survey) has picked up, that the labor force participation rate should now start rising (or at least stop falling) as discouraged workers become less discouraged (and actually start actively seeking employment) is standard (but not necessarily incorrect) economic dogma. My best guess is that the labor force participation rate has either bottomed, or is quite close to bottoming, and going forward is more likely to be either stable or rise than continue falling. And with that bit of prognostication as background, let’s move on to some hypothetical scenarios.
If the labor force participation rate does in fact rise from here, how quickly might it rise? I don’t know and neither does anybody else. Given just how unsatisfactory that answer is, I’ll invoke my economist’s prerogative and assume a scenario which at least on its surface seems reasonable enough. That assumption is that the labor force rises in the relatively near future at the same pace as which it in the relatively recent past. As can be seen in Figure 1, the participation rate was hovering more or less around 66.0% from 2003 through 2008 before it began to fall. It has taken 52 months (from October 2008 to February 2013) for the labor force participation rate to decline by 250 basis points (from 66% to 63.5%). In the rising participation rate scenarios depicted in Figure 3 I assume that it takes 52 months for the participation rate to increase by a like amount, reaching 66.0% in June 2017, after which time I assume it remains stable at 66.0%. [Note: The participation rate may well never reach 66.0% again, or at least not for a very long time, for demographic reasons. The population is aging, and the older cohort has a lower participation rate. I use the previously explained assumptions for illustrative purposes only.]
Figure 3 depicts four hypothetical scenarios, all “solving for” when the unemployment rate would reach 6.5% (I assume the population grows at a 0.97% annualized rate, by the way):
1. Job growth, as measured by the Household Survey, of 170,000 per month (the increase recorded in February 2013) with a stable labor force participation rate. The unemployment rate reaches 6.5% in March 2016. In this scenario the Fed would probably keep the fed funds rate right where it is until late 2015 or early 2016.
2. Job growth of 170,000 per month with a rising labor force participation rate. The unemployment rate rises to 9.6% in 2017 and then gradually falls but is still at roughly where it is today in a decade’s time (7.9% in February 2023). The Fed would probably provide yet more stimulus (increase QE, experiment with lowering the IOER, etc.) in such a scenario.
3. Job growth of 250,000 per month with a stable participation rate. The unemployment rate reaches 6.5% in April 2014. In this scenario the Fed would very likely start tightening sooner than mid-2015 time period that was suggested when the FOMC was still providing date-based forward guidance.
4. Job growth of 250,000 per month with a rising labor force participation rate. The unemployment rate reaches 6.5% in February 2018. This would be a tricky scenario for the Fed. Job growth would be robust, but the unemployment rate would be tediously slow to decline. My best guess is that the Fed would view the rising participation rate as an indication of just how many people had previously given up hope (but still wanted a job) and end up keeping monetary policy looser for longer, as long as inflation (and inflation expectations remained) tame. In this scenario the Fed would probably end up raising rates before the unemployment rate actually reached 6.5%, but would keep the fed funds rate where it is past mid-2015.
All of the above refers to when the FOMC might raise the fed funds rate, not when the FOMC might either stop, or reduce the size of, its monthly purchases of Treasuries and MBS (I’ll address that subject another time.)

As is illustrated in Figure 2, break-adjusted loans and leases on the books of commercial banks operating in the U.S. increased at a 6.3% annualized rate in December 2012 and at a 5.9% annualized rate in January 2013, the best 2-month string of loan growth since October and November 2011. This, of course, begs the question of why loan growth, itself, has taken a turn for the better. In true overdetermined fashion I would posit that many factors could account for better loan growth of late, including: QE3 and the creation of yet more excess reserves, the improvement in banks’ capital position, improving asset values (better loan collateral), and perhaps even the announcement (in early January of this year) of a relaxation in the Basel III Liquidity Coverage Ratio rules and implementation schedule.
From a monetarist perspective (MV = PQ, and all that), while desirable, loan growth is not an end in itself but rather a means to an end: Money supply growth. On that score I direct the readers’ attention to Figure 3 which illustrates the Center for Financial Stability’s broadest measure of money, Divisia M4, which has of late been showing some real signs of life (+6.80% year-over-year as of December 2012 and +6.13% year-over-year as of January 2013).
Will the better economic data and stock market behavior last? There are certainly downside risks (fiscal austerity has come to the U.S. and further shocks could emanate from Europe at any time) but the improvement in these monetary measures suggests that the Fed may no longer be pushing on a string. To the extent that loan and money growth are in fact responsible for the better economic data and the better equity market behavior (and I would posit that all these factors are mutually reinforcing in an overdetermined manor, i.e., that the arrows of causality run in multiple directions) I would think the answer is yes, it will last, as the Fed appears committed to keep pushing until (the still substantially negative) real GDP output gap is well on its way to having closed.

I’ve included the graph in Figure 2 simply to help the reader understand Figure 3. The simple blue line in Figure 2 represents the actual quarterly average of the SPX through 4Q 2012. The blue line with squares represents hypotheticals: The first blue square is 1479.8, which is the average closing level of the SPX thus far in 1Q 2013; the hypothetical assumption is that the 1Q 2013 average does in fact turn out to be 1479.8 (it would be higher if the SPX can hold onto yesterday’s gains for the rest of March). The second blue square assumes that yesterday’s close (1539.79) turns out to be the actual quarterly average in 2Q 2013. The simple red line is the quarterly model value of the SPX based on actual NIPA profit data through 3Q 2012. The three red dots represent the model value of the SPX based on the CBO’s estimates for NIPA profits in 4Q 2012, 1Q 2013, and 2Q 2013. As can be seen in Figure 2, there’s been a pretty wide gap in recent years between the model value of the SPX (high) and the actual value of the SPX (low), but given the hypothetical assumptions, the gap narrows significantly in 1Q 2013 and 2Q 2013.
Figure 3 takes the data (both actual and hypothetical) shown in Figure 2 and expresses this data as the actual quarterly average of the SPX as a percentage of the model value. The simple blue line in Figure 3 illustrates the relative valuation using actual data through 3Q 2012. As can be seen, this line has been meandering and trendless (below 100%) since the steep decline in 1Q 2009. However, extending this line with the assumed / hypothetical data (for 4Q 2012, 1Q 2013, and 2Q 2013) suggests that the relative valuation line is forming a rounding bottom, suggesting that the actual level of the SPX may well continue to “hone in” on the model level until the relative valuation level “reverts to the mean” and reaches 100% of the model value.
I have to admit that with the Italian elections, the messiness that will almost certainly accompany the negotiations for a Cypriot bailout, and continued uncertainty vis-a-vis the fiscal situation in the U.S. I have been skeptical that the SPX would be able to break above relative resistance at ≈86.6% of the model value in the near-term (and it’s still possible that yesterday’s close at 87.16% will prove to be just another, albeit slightly higher, relative valuation peak). That having been said, my thesis since the big decline in 1Q 2009 has been that, in the nebulous “fullness of time”, the SPX would eventually return to the model value. Assuming the NIPA profit projections I‘m using are correct, if the SPX can stay at or above yesterday’s closing level I’ll have something that I haven’t had before: A graph that at least suggests that the S&P 500 may have in fact started to revert back toward the 100% of model value mean.

Figure 2 is a summary of growth rates of “actual” real DPI (the unadjusted data as it appears in the BEA’s National Income and Product Account table 2.6) and the 2 estimated alternative measures described above. I’ll note here that it seems to me that there may still be some pre-fiscal-cliff maneuvers embedded in the November 2012 data that have not been accounted for by the aforementioned adjustments (and there may still be echoes of this in the December and January data). With that caveat, of the three measures I’d favor the last one – real DPI adjusted for all the special factors except for the expiration of the payroll tax holiday. Per this data series the presumed underlying pace of real DPI growth in recent months actually looks pretty decent with a +3.43% annualized rate of growth in over the past 3 months (between October 2012 and January 2013).
Before you get too comfortable that all is well on the real income growth front, however, please be aware of the following: Over the past three months the nominal annualized growth rate of this data series has been +2.82%. The real growth rate has been higher, at +3.43%, thanks to (what looks to be a brief) spell of outright deflation; As of January the 3-month annualized rate of PCE price index inflation was -0.59%. The rise in gasoline prices since mid-January will almost certainly put an end to this spell of deflation (and its boost to the real income growth measures) once the February data is released.

Dividend Income: Given the uncertainty surrounding the fiscal cliff and what future tax rates might be, a lot of companies made accelerated and special dividend payments in late 2012. The annualized rate of personal dividend income increased from $748.7 billion as of October 2012 to $1.05 trillion as of December 2012. Most of this late 2012 increase will probably be reversed in the January 2013 data. I calculate that if, in January 2013, the annualized rate of personal dividend income reverted back to the October 2012 level, this line item alone would act as a -2% drag on the month-over-month change in total personal income in January 2013. My best guess is that dividend income probably won’t fall all of the way back to October 2012 levels, but this line item will almost certainly be a big (I’m expecting it to be the biggest) drag on total personal income growth in January.
The End of the Payroll Tax Holiday: We’ve all seen this coming, but it should actually show up in the data starting this morning. The 2% payroll tax holiday expired at the end of 2012. This should manifest in this morning’s January 2013 income report as an increase (I estimate +$120 billion, annualized) in the “contributions for social insurance” line item (this line item subtracts from top-line personal income). If in fact this line item increased by $120 billion in January, I calculate that it would have been a 0.86% drag on month-over-month total personal income growth in January 2013. This is a “real” drag on personal income and not, as is the case with dividend income (above) and potentially private sector wage and salary income (below), just a reversal of a prior boost.
Private Wage and Salary Income: The BEA’s December 2012 Personal Income and Outlays report noted that in addition to the dividend effect, personal income in November and December, “was boosted…by accelerated bonus payments and other irregular pay in private wages and salaries in anticipation of changes in individual income tax rates.” Per the BEA, private wages and salaries were boosted by $30 billion (annualized) in December and by $15 billion in November. Given that aggregate weekly payrolls (data from the January employment report) increased by 0.32% in January 2013, my normal expectation would be that private wage and salary income increased by a similar percentage in January 2013. If so, that would equate to +$18.5 billion increase in annualized dollar terms. However, January 2013 is not a normal month for the personal income data, and reversals of the late 2012 irregularities could potentially more than offset the normal increase in wage and salary income resulting from the increase in private sector payrolls and average hourly earnings recorded for January.

I deem the “appropriate” level of the home price index to be a level which renders ratios to median family income and the owner’s equivalent rent index that are the same as (or close to) the average ratio between 1Q 1987 (when the S&P / Case-Shiller data begins) and 1Q 2001 (before the housing bubble began). As is illustrated in Figure 1 having previously been overvalued, relative to median family income home prices finished “reverting to the mean” as of 4Q 2011, and as of 4Q 2012 were actually 7.76% “overvalued” relative the 1987 – 2001 mean. While the model says home prices are now “overvalued” relative to median family income, I note that it may well be the case that home prices are about where they should be and that median family income, which has increased at a snail’s +0.2% annualized pace over the past 3 years, is in fact “undervalued”. Relative to the owners’ equivalent rent index (which has increased at a +1.31% annualized pace over the past 3 years) home prices dipped to 99.78% of the mean in 3Q 2011 and as of 4Q 2012 were 2.78% above the 1987 – 2001 mean.
The Bottom Line: As of 4Q 2012 home prices were probably just about where they “should” be.
Housing Market Activity (Sales): Note: Due to the limited historical data on co-op and condo sales, my analysis of sales refers only to the single family home sales.
How does one answer the question of what constitutes an “appropriate” pace of single family home sales? I haven’t been able to think of any better way than looking, over long periods of time, at the pace of single family home sales relative to the size of the presumed prime home-buying segment of the population, which I define as the population of 25 through 54-year-olds. I deem the “appropriate” pace of single family home sales as being a pace that produces a ratio of sales to population that is equal to the average ratio from December 1983 through December 2001.
As is illustrated in Figure 2, as of January 2013 the pace of total single family home sales was 96.9% of “normal. The pace of existing single family sales was 105.5% of the model value as of January 2013 (the pace of single family sales has been above 100% of the model value since August 2012). While showing a notable improvement after January’s +15.6% month-over-month increase, the odd man out here continues to be new home sales, the pace of which as of January 2013 was still only 53.5% of the model value.
New Home Inventory: So why is the pace of new home sales still so far below the presumed “appropriate” pace? In my view the most pertinent explanatory factor is the low level of new homes available for sale. People can’t buy what’s not up for sale. The U.S. Census Bureau’s data on new home sales begins as of January 1963. As can be seen in Figure 3, the all-time low inventory level was 143,000 in both July and August 2012. As of both December 2012 and January 2013 the number of new homes for sale had increased by a bit, but only by a bit, to 150,000. As of January 2013 the months of supply metric was just 4.1 months (and that was with the pace of new home sales running at about half what would be considered “normal”).
My Expectation: Lots of new homes are going to be built this year. As the supply of new homes increases, so will the pace of new home sales.