North America Cracks of driving liquidity

5 min
Alexis Garatti
Alexis Garatti Head of Macroeconomic Research
Dan North
Dan North Senior Economist for North America

A few segments of the US economy show some signs of fragility amid higher rates

 

A significant increase of the Libor rate

The recent announcements made by the White House on trade protectionism have triggered a sudden surge in the level of uncertainty, harshly penalizing equity markets and increasing the demand for safe haven assets. As a result of this, US long-term interest rates have markedly declined over the last month from 1.9% to 1.8%. In the opposite way, and despite rising uncertainties, US short-term interest rates have continued increasing, in particular the LIBOR rate, which today reaches 2.35%. This trend reflects above all the confidence of the market in the Fed’s capacity to continue its monetary policy normalization with a gradual increase of official rates. We still expect two more hikes in 2018.

More worryingly, the LIBOR – OIS spread currently reveals increasing difficulties of banks in accessing short-term liquidities on the money market. During the subprime crisis, this indicator played the role of a bell-weather on the health of US banks as sudden reversal of confidence in any financial conditions could lead to a market freezing, with strongly negative impacts on banks’ credit conditions and on the real side of the economy. Today, this indicator has lost of its relevancy because of regulatory changes, which have transformed the way money market funds interact with banks. Despite this new element, the widening of the LIBOR – OIS spread still contains some information in terms of liquidity conditions, as evidenced by our proprietary indicator of excess supply of money.

In order to identify the impact of the Fed’s tightening policy (progressive increase of official rates and reduction of the Fed’s balance sheet), we estimate a theoretical money demand based on M0 and quantity theory of money, i.e. we calculate a theoretical M0 (transactional money demand), which is determined by the relation MV = PT with M corresponding to M0, V indicating the velocity of M0, P the level of prices and T the level of transactions or nominal GDP. We study this relation in variation and compare the observed evolution of M0 and the theoretical one justified by activity and speed of money circulation. To this regard, it appears that we are currently at an interesting juncture, where the excess of liquidity is now close to 0 after long years of money excess supply, during which debt and therefore investment financing were particularly cheap. Our indicators therefore points toward an end of the cheap money, which could have a significant impact on the real side of the company, for those economic actors which rely the most on debt.

The high yield sector is probably one of the most exposed segments of the US economy to an interest rate shocks as it benefitted from exceptionally advantageous interest rate conditions. On the basis of a bottom up process using financial data from companies listed in the S&P500, we can have a view on the relative level of leverage of sectors calculated as the ratio of net debt to ebitda. To this regard, the utilities sector is the most leveraged, followed by materials and energy, confirming the existence of high indebtedness in sectors which are relatively more capital intensive.

We study the sensitiveness of CDS spreads by sector to LIBOR rates and find that indeed a context of higher rates leads to an increase of the probability of default in only two sectors, i.e. consumer discretionary and utilities, while energy sector is likely to suffer as well. Our data basis confirms than energy and retail registered the highest number of major insolvencies (with services, which is a collection of disparate activities), respectively 5 and 16 in 2017.

REIT prices falling

Because of their exposure to retail activities and their sensitiveness to rates, REITs (Real Estate Investment Trusts) could be exposed to interest rate shocks. They are similar to mutual funds although they invest primarily in commercial real estate.  REITs yield an income stream from the rents on the properties. This income stream, which is paid in the form of a dividend, is highly correlated with LIBOR. Dividends tend to rise as LIBOR rises.  Over the past six months, LIBOR has risen by approximately +80 basis points (0.8%) while REIT prices have fallen -6.2%.  The market value of all REITs is approximately $1T, whereas the market value of the S&P 500 index is approximately $25T.

Delinquencies on Consumer Debt Rising

Student, auto and credit card debt

Consumer debt has been growing at an average rate of 4.4% y/y since 2003, outpacing nominal GDP growth of about 4.0% y/y over the same period. This faster growth rate implies that the economy is now somewhat more exposed to consumer debt than before. Credit card debt has grown on average 1.3% y/y since 2003, auto loans have grown 4.0% annually, but student loan debt has soared at an annual rate of 12.9% from $0.25T to $1.4T in just 14 years. More to the point, despite the fact that the outstanding debts are similar in magnitude to student debt, the delinquency rate is something completely different.  Chart 5 a, b, c shows the amount of debt outstanding which is delinquent by 30 days. Clearly student debt is by far the worst performer with an 8% delinquency rate, resulting in $106B of delinquent debt outstanding. The credit card debt delinquency rate is only 2.5%, although it is worth noting that the amount of delinquent debt has risen sharply over the past four years from $16B to $20B. Auto loans have a delinquency rate of only 1%, but again have almost doubled in four years from $10B to $18B.   And while delinquencies are increasing in all three categories of consumer debt, they are all exposed to variations in short-term interest rates as well. Credit card rates are perhaps the most sensitive to LIBOR.  And as LIBOR has risen around 0.85% over the past year, credit card rates have used the excuse to skyrocket about 1.5% to almost 15.5%.

Chart 5 a, b, c: Consumer rates and LIBOR

 

Sources: Bloomberg, Euler Hermes, Allianz

Mortgages

Residential mortgages are by far the largest portion of total consumer debt. Currently there is $12T of consumer, residential mortgage debt outstanding, which is about three quarters of the total $16T in consumer debt outstanding.  Mortgage activity is highly sensitive to LIBOR rates as shown in Chart 6; the LIBOR rate is shown inverted demonstrating that when rates go up (the red line goes down), mortgage applications go down. Mortgage refinancings are particularly sensitive to LIBOR variations.

An increase in LIBOR will raise mortgage rates, increasing the monthly payment borrowers must make, effectively shutting some borrowers out of the market. The median price of an existing home over the last year was approximately $250,000. An increase in the 30 year fixed rate from 4% to 5% would increase the monthly payment from $1,194 to $1,342, or a sharp 12.4%. The median price of a new home (the new home market is about one-tenth the size of the existing home market) has been about $325,000.  That same increase in the 30 year fixed rate from 4% to 5% would increase the monthly payment from $1,552 to $1,745, again a sharp 12.4% increase.  Clearly monthly mortgage payments are very sensitive to movements in the 30 year fixed rate, which is highly correlated to LIBOR. Rising LIBOR rates will have a significant negative impact on the mortgage market.

In addition to rising rates, there are other signs of stress in the mortgage market as well. First, median home prices have been rising much faster than median incomes as shown by the tan line in chart 7. In fact since 2012 the price of a median existing home has risen at an average annual rate of 7.7%, whereas the median family income has risen at less than half that rate, 3.1%. Prices have been driven up by a lack of homes for sale and a strong job market. 

Chart 6  a, b: loans demand and Libor

Sources: Bloomberg, Euler Hermes, Allianz

As a result, buyers are having to stretch for houses, and their monthly payment as a percentage of their income is rising sharply as well. And while the payment/income ratio has risen from about 2.5 to 3.5, that ratio will continue to go up with rising interest rates; the data shown here only goes through January and does not reflect the rapid rise in LIBOR since then.

This lack of affordability has caused participants in the mortgage market to lower standards to make it easier for potential buyers to get a mortgage.  While the goal of making housing more accessible may be a virtuous one, it does create increased risks in the mortgage market.

Chart 7 Declining affordability

Sources: Bloomberg, Euler Hermes, Allianz

For example, the two largest participants in the mortgage market, the government sponsored entities (GSEs) Fannie Mae and Freddy Mac recently raised the allowable debt-to-income ratio from 45% to 50%. The result according to CoreLogic is that around 20% of mortgages recently issued had debt-to-income ratios greater than 45%, three times as much as in 2016 and early 2017. And the Urban Institute reports that the share of mortgages which had both higher debt-to-income ratios and lower credit scores rose from 19% a year ago to 25% in the first two months of this year.

Other risky practices are emerging. Fannie and Freddie are now backing lenders who will help pay a borrowing student, and guaranteeing loans with down payments as low as 3% - the standard used to be 20%. And one of the most dangerous products which sparked the housing meltdown of the 2000s, sub-prime loans, has returned with a new name: non-prime loans.

These loans are being made by “non-bank” lenders who are in fact getting their funds from the largest U.S. banks such as Wells Fargo, Citigroup, and virtually all of the other major banks who needed to be bailed out from the risks of this very type of loan in the 2000s. Loans to non-bank firms are now among the largest items on some banks’ balance sheets. Remarkably some of these loans are being made on very poor risks with histories of foreclosures, bankruptcies, bad payment, and credit scores as low as 500. And they are being issued in amounts of up to $1.5 million, including cash outs of $500 million. Disturbingly, for the first time in 30 years, the majority of mortgage loans in dollars are being made by non-bank lenders.

Clearly, demand in the mortgage markets is very high, and lenders are taking on imprudent levels of risk just ten years after the mortgage meltdown of the 2000s. With the higher risks on outstanding mortgages, an increase in mortgage rates based on LIBOR could put a significant strain on the market.  As rates rise, fewer potential borrowers will qualify for mortgages, but lenders seem likely to keep lowering standards.  Adjustable Rate Mortgages (ARMs) could see a significant, even damaging, increase in defaults. The era of improving credit quality seems to be over in the US.