February 2006

'Neff Said?

'Neff Said...In the wonderful first edition of the Barron's annual roundtable discussions that hit the Street a few weeks back, we saw a quote by John Neff that pretty much stood out like a sore thumb to us.  And the reason we bring this up is that John's comments sure seem to be reflective of broader consensus thinking these days regarding the US consumer, the general dismissal of the US savings rate as having any real meaning in the current environment, and very reflective of what the consensus expects for the macro US economy.  Before taking even one step further, we'll be the first to admit that we have a ton of respect for Neff.  He has clearly proven himself as one of the better investors of our time.  In fact, that's exactly why we always make it a point to listen to what he has to say.  If we can approach even one half the investment success John has had over time, we'd consider it a minor miracle.  Our singling out Neff has nothing to do with Neff, per se, but rather the line of reasoning that seems all to easy today.  Here are the appropriate quotes taken from the roundtable discussion verbatim:

As you know, there are a number of messages in John's comments.  First, he's implying that the savings rate calculation is either flawed or inappropriate for characterizing the current period.  And believe us, he's wildly far from alone amongst the Street pundit community.  Secondly, he's absolutely correct in pointing out that indeed households are flush with cash/near liquid holdings bordering on the $5 trillion number.  Any way you look at it that's a lot of liquidity and is a record number from a historical standpoint, if we're not incorrect.  Neff is suggesting that the US consumer is plenty liquid and more than able to continue spending, especially given the psychological boost of heightened equity and real estate prices.  OK, fair enough.  We're not disagreeing with Neff's comments in absolute terms.  He's dead on.  But where we believe Neff should have continued, and where we believe most pundits end this convenient story of the moment, is in not pointing out how these facts presented in isolation fit into the context of broader household financial circumstances of the moment.   It's somewhat like us telling you we have a home worth $1 million, but forgetting to add that we have a $900,000 mortgage against it.  

We believe that one of the most important exercises in contrarian thinking is trying to specifically analyze and characterize a stream of logic that is both widely held among market participants and works to support their current investment actions, but is inherently flawed.  The second step in the process is to bet against it in what we can only hope is well timed fashion.  As you know, much easier said than done, now isn't it?  Especially the timing part.

We promise to move through this in relatively quick fashion.  Bullet point arguments and a good measure of "pictures" to support the thinking.  To be honest, we're a bit surprised at Neff's savings rate comments.  Why?  Because the US savings rate calculation being presented by the BEA (Bureau of Economic Analysis) each month in the personal income numbers report has been a consistent formulaic calculation for over a half century now.  No adjusting.  No changing the calculation for political or perceptual reasons.  In other words, unlike so many government stats of the moment, it's directly comparable across history.  In our minds, it's one of the beauties, if not the single most important characteristic of this data series.

First, a total long term picture of the US savings rate calculation as per the historical BEA data.  You've seen this before.  No huge surprises here.  The drop off begins as the current US credit cycle begins to grow much deeper roots in the 1980's.  And from there, the rest is history, so to speak.  And yes, this drop off period coincides almost perfectly with Neff's comments of stock and real estate values ascending skyward, but not being counted in the official savings rate in terms of these being "unrealized gains".
       

What we thought we'd do in the next four charts is to look back across history to see if there have been any other periods of meaningful growth in residential real estate and common stock prices that rival the current experience on a rate of change basis.  If so, let's see what happened to the US savings rate during those prior periods of very meaningful price acceleration.  If indeed acceleration and deceleration in home and common stock prices have influenced consumer behavior in the past, as Neff is implying is occurring at the present in terms of substituting stock and home price gains for the actual act of saving, we'd expect to at least see some big dips along the way in the US savings rate as real estate and common stock prices spiked on a rate of change basis over time.  Very simply, is there a savings substitution effect or isn't there?  Yes or no.  Pretty simply stuff, right?  What lies below is probably all too familiar to you.  It's the OFHEO (Office of Federal Housing Oversight - the GSE regulator) rendition of macro year over year change in US residential real estate prices.  We've marked the prior relatively meaningful rate of change peaks.  As you can see, the early 1978 experience even bested what we have just lived through in the current cycle.
        

If indeed households are so sensitive in their spending and saving patterns to changes in housing prices, why the heck was the US savings rate 9.1% in 1978, the period of the greatest year over year change in US residential real estate prices in literally the last three decades?  Beats us.  We approach this same question just a bit differently in the chart below.  What we've done is gone back and looked directly at household balance sheets.  We're looking directly at the value of actual household real estate holdings as opposed to more of a generic price index as seen in the OFHEO data above.  Now certainly this data encompasses more than just price.  It also encompasses the effect of purchases in terms of increasing year over year holdings.  But, let's face it, do households buy real estate if they are not confident in the asset as an investment?  Just look at what's happened in the last three to five years for an answer to that question.  Balance sheet values importantly reflect confidence as well as price.  Once again, you can see the peaks in terms of rate of change.
       

Let's follow up with the obligatory recantation of historical savings rate levels associated with these rate of change price peaks.

Funny, household real estate holdings skyrocket in the 1970's and 80's with growth rates similar or greater to what we've experienced in the current cycle, and yet still the savings rates in the 70's and 80's remained at or near double digit numbers.  So if changes in housing prices don't seem to have influenced the US savings rate, it must be stock prices, right?  Let's see.  To try to get a sense for how near term stock price returns affect subsequent US household saving rate activity, we've constructed the chart below that chronicles the price only three year moving rate of return for the S&P 500, as a proxy for stock price returns in the aggregate.  We've chosen three years because we believe relatively near term experiences influence household decisions, and more importantly, the current US equity market rally has been rolling for about three years now.  If Neff believes stocks are acting to support consumer well being now, reflected in the apparent lack of need to save, we've only really recently experienced three years of good numbers.  We're trying to give Neff's line of reasoning the total benefit of the doubt here.  You can clearly see the dates of prior three year moving average stock price peaks below.  Good times, right?
        

And as you'd expect, once again here come the savings rate historical comparatives.  Let's see if stocks will do the trick and work the lack of need for actual savings magic to which Neff seems to be referring.

Close, but no cigar, huh?  Historically, even very meaningful and extended periods (three years) of stock price gains have not been enough to dissuade households from actually putting a few dollars in the piggybank for a rainy day.  This was even true seven short years ago in 1998.  One last chart below, and this time we'll spare you the historical savings rate comparatives because the message simply reinforces the examples we have already been through.  Like stock market price change data above, we're looking at the three year change in household ownership of common stocks with data directly from the Fed themselves.  Again, this encompasses not only price appreciation, but includes the influence of additional purchases or sales.  And we're convinced the purchases or sales portion of the numbers importantly reflect confidence, which we would hope is also translated into offsetting savings rate activity.
       

Interesting.  Clearly there have been times over the last five plus decades where the three year moving average rate of change in household ownership of equities was well above what we now experience.  Should we not have seen meaningful dips in the US savings rate as this true multiplicity of price peaks was seen again and again over time?  Shouldn't we have at least seen savings rate dips when both household real estate and stock price peaks coincided, or roughly coincided throughout history?  That's the logic that's being applied to the dismissal of meaning in today's savings rate calculation.  So just why shouldn't that also have been true of the past?  Clearly, implicit in Neff and many a pundits comments of today, whether they realize it or not, is that "it's apparently very, very different this time" when it comes to interpreting the message implicit in today's negative savings rate.  Again, for ourselves as investors, is this interpretation right or wrong?  Of course the correct answer for the current cycle lies ahead.  But with absolutely no historical backing for this belief that the negative US savings rate is to be ignored at present due to significant home and equity price gains of the recent past, and the relatively widespread current mainstream conviction that the savings rate should be ignored, we'll take the other side of the consensus bet, thank you.

Of course we can't yet end this discussion without at least a comment or two regarding the $5 trillion in cash at the household level to which Neff refers.  We've been hearing the "mountain of money" argument for at least a good decade now.  On face value we have two visceral comments.  If the mountain of money is so powerful, why does it keep growing in nominal terms and why hasn't it acted to explode the financial markets upward?  Secondly, if there is so much money around, why did the homebuilding and housing finance industry ever have to resort to 0% down or exotic option ARM lending practices?  Moreover, with so much money sitting on household balance sheets, why have households taken on so much mortgage debt over the past five years, especially when that "cash" was earnings so little in terms of rate of return during Greenspan's 1% Fed Funds rate shock and awe campaign?  Why didn't they use a good portion of their existing cash to "invest" in real estate?  We suggest quite humbly that these are very simplistic and commonsensical questions.  No incredible insights or rocket science from us.  But as you would imagine, we personally choose to focus on the $5 trillion in cash in a broader context.  And that broader context is quite simply to look at both sides of the household balance sheet, not just the left side singularly.  Yes, cash has grown, but what has happened to household liabilities over there on the right side of the household balance sheet to offset this "compensating balance" called cash?

In the following look at the historical structure of the household balance sheet, we're going to one up Neff in what we think is a pretty big way, with the thought of giving households more than the total benefit of the doubt.  If one looks at total household liquidity to include not only cash (savings, CD's, checking, MMF's, etc.), but also to include every single nickel of household holdings of bonds (USTs, corporates, muni's, etc.) and assume these bonds could be sold in a heartbeat without any price pressure at all, the real number for total household access to "liquidity" at the end of the third quarter of last year is $8.2 trillion, not a measly $5 trillion.  Now we're talkin' baby.  OK, let's use the definition of household liquidity we've described - all cash and all bond assets - in the charts below.  First, here's the nominal dollar history of household liquidity and liabilities over the last six decades.
      

Now that's interesting.  From 1945 up until 1995, household cash (again, using our very broad definition) always exceeded household liabilities.  It's really only been in the last ten years where household liability growth has shot up like a bottle rocket to far exceed household liquidity (cash and bond holdings).  Another very simple way of looking at this same set of numbers is simply to subtract household liabilities from our broad definition of household cash/liquidity.  And here you have it.
        

The chart above does spark yet a few more commonsensical questions from our rather simple minds.  As opposed to engaging in yet more consumption, why wouldn't households choose to use some of their apparently very ample nominal dollar cash to reconcile what you see above?  Why wouldn't they pay down some debt in an attempt to normalize the relationship you see above to nearer historic averages than not?  Again, especially because yield on cash and bonds in the past few years has been terrible.  And what would happen to these numbers if indeed consumers took Neff's $5 trillion in apparently spendable cash and did just that - spend it on consumption?  As you know, had we used the $5 trillion cash number as opposed to our enhanced definition of household cash at $8.2 trillion, all of the charts above would look much different and much darker than they do.

One last chart that is again a bit redundant at this juncture, but we hope makes more of a social phenomenon or group think point than not.  It's simply household cash as a percentage of household liabilities (again, using our enhanced definition of liquidity that currently stands at $8.2 trillion).  It's a clear fact that back in the early 1950's, many folks had the depression and its financial effects on their families very fresh in their minds.  Having watched their indebted neighbors or loved ones basically financially wiped out, the aversion to leverage was incredible.  In the early 1950's, in aggregate, household liabilities could have been extinguished with available household total liquidity in a ratio of over two and one half to one.  Slowly as the horrid memories of the depression faded, so did the need for families to hold so much liquidity.  But the change in thinking since the mid-1980's, as is reflected in this chart, is more than striking.  Certainly this accelerated as US asset price/monetary inflation exploded in the last few decades.  At the end of 3Q 2005, household cash/liquidity as a percentage of liabilities stood at the lowest number on record in this Fed sponsored data series.  And of course at present, we have just appointed a Fed Chairman who sincerely believes the 1930's depression could have been avoided entirely had the Fed simply printed enormous amounts of money.  Of course implicit in that thinking is that the depression could have been avoided had households just borrowed enormous sums of this newly printed money.  (It's just a shame that Bernanke does not follow through in his academic conviction and tell us just how the banks at the time would have been more than happy to lend huge sums of money into an economy characterized by a 25% unemployment rate.)  It's our view that the chart below chronicles polar opposites in terms of societal acceptance of leverage.
       

Again, we have not gone through this little exercise to in any way knock the intelligence or insight of John Neff.  We suggest he deserves more than a good deal of respect given his experience in the financial markets.  We just think Neff's quick comments, as well as many a modern day mainstream pundit echoing the same line of reasoning, are very reflective of consensus thinking and deserve to be challenged.  The outcome to this little savings rate controversy certainly lies ahead.  For now the endgame is an unknown.  But history is "telling us" not to ignore the message of a negative savings rate at present.  It's also telling us to view household financial circumstances in their entirety, not dwelling specifically or uniquely on either side of the balance sheet at the exclusion of the other.  Before we let you go, we'll pull one of those old market folklore comments out of the treasure vault.  "Imbalance or abnormality is never so dangerous as when it is widely perceived or accepted as being normal."   'Neff said?  Yeah, we think so.