Thursday, January 29, 2009

Reexamining Monetary Cause and Effect

Last November, I wrote an extensive article that raised some major questions about how our monetary/credit system really works. I circulated it among some economists and market professionals. Some found it very insightful, while others thought it was crazy.
One way or the other, I think it raises some very key points that you won't read anywhere else, so I am adapting its best elements to this blog. I have already discussed many, but not all, of the themes.

Hypothesis:
The rise of capital markets, foreign exchange and global trade has rendered traditional monetary tools ineffective, and at times, counterproductive.
Key themes:
• The last Fed tightening cycle may have increased liquidity rather than reducing it.
• New sources of credit have risen for the U.S. economy that established monetary policy does not contemplate: Foreigners investors, securitization and currency markets.
• Banks have declined in importance as lenders to the economy.
• Determined not to copy the Fed’s mistakes of the 1930s, policymakers are erroneously confident in low interest rates and fiscal stimulus as a means to solve the current crisis.
• The most effective solution may be to restore demand for spread-based credit products. One possible response would be the creation of a U.S. sovereign wealth fund, which could have the added long-term benefit of helping pay for Social Security and Medicare.

OBSERVATION #1:
Funding entities such as Structured Investment Vehicles (SIV) channeled short-term money into long-term asset classes, undermining the Fed’s attempts to restrain credit growth in the 2004-2007 period. In the recent credit bubble, SIVs borrowed in the short-term by issuing commercial paper and invested their proceeds in longer dated credit instruments, mainly asset-backed securities (ABS) linked to home mortgages.
About six months after the Fed started raising interest rates in June 2004, money started flowing into asset-backed commercial paper (ABCP). The Fed steadily raised its rate for the next two years, from an original 1% to 5.25%. During that period, the amount of ABCP outstanding would grow $531bln, or about 80%, to $1.18 trillion.
Importantly, this amount peaked in July 2007 at the same time short-term interest rates reached their high. Three-month Treasury bills yielded 4.83 on July 20, 2007. As pressure mounted for the Federal Reserve to cut interest rates in August 2007, short-term yields dropped. This appears to have exacerbated the selling of ABCP, which was already under pressure as investors worried about its link to mortgage debt.
Between August 8, 2007 and October 15, 2008, the amount of ABCP fell $534bln, or 44%, to about $676bln. About $531bln flowed in as the Fed raised rates, and another $534bln left as the Fed cut interest rates back to 1%. This leveraging process matches the Fed’s interest rate moves closely.

The approximate $530bln figure represents 3.86% of 2007 GDP, 4.78% of total mortgage loans in the economy, and about 14% of ABS outstanding. Liquidating this large amount represented the opening round in the credit crunch.

The set of charts below highlights the correlation between higher rates and money in ABCP. I have never seen anyone else put this data together and think it merits further discussion.

SIVs would have unwound regardless of the Fed’s rate cuts. The important point is their structure caused them to directly oppose the Fed’s policy goals: They attracted money when short-term rates rose and lost money when rates fell. Because the same money was then recycled into credit securities, higher short-term rates had the unintended consequence of increasing risk appetite rather than reducing it.
Similarly, an inverted yield curve caused some investors to seek even higher yields to cover their own rising borrowing costs. Hedge funds, for instance, could borrow at a Libor-based rate when Libor was 2-3% and long-term corporates and ABS yielded 5-7%. (They probably borrowed at about Libor+100bp. The banks lending to them, like Goldman Sachs, were simultaneously funding themselves at about Libor+10-40bp.)
As Libor moved towards 5%, the universe of profitable investments shrank. Hedge funds responded by purchasing lower-quality assets, fueling huge growth for collateralized debt obligations (CDO) and leveraged loans. They had a similar effect as the SIVs, boosting demand for long-term (3yr+) credit securities as short-term rates rose.

The key message once again is that higher rates apparently encouraged, rather than discouraged, risk appetite. This precisely matched my own anecdotal experience as a reporter covering the bond market. Given this history, I am worried by the insistence of almost everyone in the market that lower rates are the answer.
OBSERVATION #2:
Currency-based financing tools, such as the yen carry trade, had similar effects as SIVs and hedge funds -- probably on a larger scale.
The Fed’s constant interest-rate increases corresponded to a steady appreciation of the dollar against the yen. This allowed market participants to borrow yen at very low rates and to reinvest in higher yielding assets. (Many times it was reinvested in currencies other than the U.S. dollar.) The only risk in the carry trade is when the yen rises quickly, which increases the size of the investor’s liability. This inverse relationship between the yen and “risky” assets has been well established in recent years: Almost every day that stocks fall, the yen rallies.
The Fed gradually increased its rate from 1% to 5.25% between June 2004 and June 2006. Because the tightening was gradual and predictable, the dollar’s appreciation versus the yen was also steady, making investors comfortable selling yen. The dollar gained steadily against the yen as the credit bubble inflated between the start of 2005 and the middle of 2007. In today’s market, dollar strength versus the yen is equivalent to cutting interest rates because it facilitates borrowing, and vice versa.

OBSERVATION #3:
Direct foreign capital flows into the U.S. made an unexpected and powerful contribution to asset-price appreciation.
Between June 1998 and May 2007, the U.S. credit market experienced a significant inflow of foreign capital that substantially increased the availability of funds for businesses and households. This resulted from a large growth in the trade deficit, which approached 6% of GDP in 2005 and 2006, compared with less than 1.5% in the 1990s.
These funds primarily flowed into an asset class the Treasury and Federal Reserve report as “corporate bonds,” which includes traditional corporate bonds and non-GSE securitizations(ABS). I call these “U.S. credit products.” The foreign money also flowed into GSE debt.
Over the period of inflows, foreigners bought more than $2 trillion dollars of U.S. credit products, pushing them past life insurers as the biggest holders. Credit products were also briefly foreign investors’ largest asset class in the U.S., surpassing stocks in Q2 2007, according to the Fed’s Flow of Funds report. (In Q3 2008, Treasuries retook the top spot.)
It should also be observed that money behaves differently when it originates in trade. If Americans had consumed products from their own country, most of the money would have been paid to Americans in wages and taxes. Only a small proportion would have filtered through to the capital market. In contrast, when the cash originates from trade, most of it is channeled into the capital market. From 2003-2007, $1.8 trillion, or 54% of the cumulativ U.S. trade deficit, was recycled back into U.S. credit products – excluding Treasuries.
This flow of overseas money into U.S. fixed-income assets was a major factor helping push borrowing costs lower for several years. (It is probably under appreciated because it was the first time foreign capital noticeably impacted the U.S. since before WWI.)

For more on the relationship between trade and credit excesses, see this blog entry.

OBSERVATION #4:
Several sources of liquidity reached their apogee at the same time traditional corporate borrowers had the least need for funds. As discussed, these sources of liquidity were overseas investors, the yen-carry trade and short-term funding schemes such as SIVs.
Traditional corporate borrowers, such as telecom companies, had a declining need for debt capital because they had just completed a significant investment in new networks, etc. The period of the late 1990s experienced strong capital expenditure, resulting in a natural downturn afterwards.

Furthermore, early in the period of extreme capital-market liquidity, traditional mortgages lenders Fannie Mae and Freddie Mac reduced their activities due to accounting scandals. This means large amounts of money entered the market in 2004-6 at the same time traditional users of capital had much less need for funds.

The financial industry responded to this flood of money with a surge of private-label mortgage lending. Non-financial corporations reacted with record amounts of stock buybacks and acquisitions. Private-equity funds raised unprecedented amounts of capital they hoped to leverage using high-yield bonds.
The financial industry created new structured products, such as CDOs, which increased the demand for mortgage bonds. In the chart below, ABS issuance is compared with debt sales by traditional non-financial corporates. This surge captures the growth of subprime lending and highlights the true nature of the credit bubble. The key consideration is that the size of the bond market grew dramatically during the time span captured in the chart below, which makes the surge in ABS issuance even more important.

The bottom line: Excessive liquidity in the capital markets, intermediated by non-traditional
and unregulated lenders, drove home prices higher. Trade imbalances and novel sources
of leverage/funds created large pools of capital with nowhere to go. Wall Street
responded by “innovating” new products.

OBSERVATION #5:
The crisis of September and October 2008 after the Lehman Brothers bankruptcy resulted from problems in the long-term debt market (capital market) rather than the interbank market.
A timeline of events:
Sept. 15: Lehman Brothers Ch. 11 filing
Sept. 16: Eurodollar deposits (Libor) rise from 3% to 3.2%
Sept. 17: General money-market funds, which hold corporate commercial paper, report
$220bln outflows over the preceding week. Another $220bln pours from the funds over
the following month.
Sept. 17: Financial sector bond prices fall 1-3 points. Eurodollar deposits rise to 3.75%.
Stocks plunge. The Reserve Fund money market falls below $1.
Sept. 18: Eurodollars rise to 5%.
Sept. 30: Eurodollars rise to 6%
What happened? Here’s my interpretation:
• Lehman Brothers failed.
• The Reserve Fund owned Lehman CP, which was now worth less than face value. This pushed its share price below $1.
• In panic, investors pulled money from all non-government money markets.
• Money market funds were forced to raise cash to meet these redemptions. They appear to have sold large amounts of short-term financial-sector bonds (due in 1-3 years).
• Lower prices pushed yields higher for short-term financial paper, causing the sector’s curve to invert.
• Many short-term financial bonds are tied to Libor. When their yields rose, Libor was forced higher. At the same time, financial panic pushed short-term Treasury rates to historic lows. Unlike in previous downturns, this time the fate of the financial sector and the “safe” government sector diverged.
• This is why the “TED spread” gapped wider. People said it resulted from a lack of confidence between the banks. Interestingly, the rates for actual commercial paper did not move higher until later in the period of liquidation in late September 2008. (People have come to describe the TED spread as a measure of panic or risk-aversion in the market. I think it's interesting that during the several years I covered the corporate-bond market, no one ever considered it worth mentioning.)

The problem wasn’t in the interbank market, it was in short-term (1-3 year)
financial bonds. I for one never understood all this talk about the interbank market, because during most of 2007, interbank debt represented less than 0.5% of bank liabilities. Their bond-market debts were 13 times bigger. This is why the FDIC's backing of bank bonds under the Temporary Liqudity Guarantee Program (TLGP) has been so beneficial. I am not an expert in banking, but all this talk about the interbank market always seemed like a red herring to me.

This chart shows how billions flowed out of money market funds the week ending Sept. 17 as Lehman failed. It continued in following periods.

Source: AMG Data Services


Yields for short-term financial-sector bonds rose as money market funds dumped them on the market. Notice in the chart below the yield on the 1-3yr financial paper (top line) rises before
anything else and pulls the TED spread higher:

Source: Merrill Lynch, Federal Reserve

The chart below shows the unpredented size of the price decline in short-term financial sector bonds. This is what drove their yields higher and forced Libor to increase:

Source: Merrill Lynch

Here I would like to offer a hypothethical scenario to explain. Say an investor had a line of credit at Citigroup that cost L+50bp. If Libor is 3%, his cost of funds is 3.5%. Say Citigroup's has 2-yr notes trading yielded about 3.25% in early September.

Then Lehman fails and money market funds unload those same Citi bonds, forcing them to drop to 90 cents on the dollar, and pushing their yield to say 4.25%. If Libor stays the same, the investor's cost of funds is still 3.50%, giving it the ability to earn 75bp of carry trade. Under this situation, the investor would borrow as much as possible to buy high-yielding Citi notes. This increased demand for funds would naturally push Libor higher.

That kind of easy money cannot exist for long in any financial market. Given the fact that you could replace the name of Citigroup with any other major bank, it's not hard to see why the lenders themselves pushed Libor higher. Don't forget, a committee of banks in London are the folks who set Libor in the first place.

At the same time this was happening, all of that money that exited the money markets, and exited the 2-yr Citi notes, went into US Treasury bills. This forced government rates extremely low at the same time that Libor was being pushed higher.

Of course, the scenario I describe above is only hypothetical. But it describes how the gaping TED spread resulted from developments in the corporate bond market, rather than a panic in the interbank market. There was an unusual, if not utterly unprecedented, divergence between Libor and Treasury yields. But it resulted from normal market dynamics of supply and demand. Again, this was a capital-market based crisis, not a normal banking crisis. Also, if it were a normal banking crisis, the problems would have started at commercial banks, rather than securities firms. This is why I take issue with academics like Nouriel Roubini who want to apply a Swedish nationalization model to our system, which is utterly different because it's based on an Anglo-Saxon capital market model. I don't think the world has ever seen a financial crisis like this one. We need to focus on the bond market, not normal loans.

OBSERVATION #6:
The Fed Funds rate appears to have lost its efficacy starting in 2004. Using three-month
Treasury bills as a proxy for the Fed funds rate, I compare the effectiveness of various
tightening cycles:
June 1980-June 1982:
Short-term rates rose significantly (about 500bp from June 1980 to June 1982. During
that period, the growth of household borrowing slowed from 10-17% levels to 4-5%. This was the intention of Fed Chairman Paul Volcker, and it represented a successful use of the overnight rate to slow credit creation.
March 1988-March 1989:
Short-term rates started rising in March 1988 and peaked a year later. (Three-month
Treasury yields rose from 5.80% to 9.33%.)
Household borrowing growth slowed from over 9% in Q1 1988 to about 7.3% at the start of 1989 Rates peaked in March 1989 and fell for the rest of the year. This allowed household borrowing to reaccelerate, growing 11.23% in Q4 1989. Then the economy faced recession and the savings and loan crisis. Again in this case, short-term rate increases succeeded in discouraging credit growth.
March 1994-December 1994:

From March 1994-December 1994, short term rates rose from 3.33% to 5.84%.
Household borrowing fell with a small lag, dropping from 8.93% in Q4 of 1994 to 5.6% in Q4 1995. Short-term rates settled around 5% between April 1996 through September 1998, during which time household borrowing grew in the 6-8% range.
Again, it appears relatively tight monetary policy discouraged excessive household borrowing.
June 1999-May 2000:
This tightening phase was interrupted by worries about computer failures at the turn of the century. The overnight lending rate rose from 4.75 to 6.50%. As the increases were occurring, bank credit and household borrowing both increased. Soon after the increases peaked in May 2000, bank credit slowed. Household borrowing edged higher, growing in the 8-10% range quarter-on-quarter versus a 4-9% range previously. It’s hard to measure this period because it was followed by aggressive rate cuts, during which time household borrowing spiked higher.
This phase shows an unclear success of the Fed’s overnight rate in controlling credit growth in the economy.
June 2004-September 2006:
The Fed started a tightening phase in June 2004. Short-term rates would climb from 1% in May 2004 to about 5% in September 2006. They remained around 5% until July 2007. As the Fed raised rates between June 2004 and September 2006, household borrowing did not fall. It essentially remained in the 11% growth range. It appears that in this cycle, the Fed funds rate failed to restrain credit growth.

This was the great credit bubble that Alan Greenspan described as a “conundrum”. Borrowing broke loose from the Fed’s grasp. (Interestingly, and unlike before, almost all of the growth came from home mortgages rather than consumer credit.) The credit bubble formed as the Fed was raising rates.

Let’s consider rate-easing moves as well.
March 1989-July 1990:

Short-term rates fell from 9% to under 8%. No discernable improvement can be observed in bank credit, and household borrowing declines.
November 2000-February 2002:
Short-term rates fall from 6.36% to 1.76%. Household borrowing growth rises from about 8% to about 9%. This was a modest improvement, and was countered by the recession and stock market crash.
The Fed cuts again in June 2002 and rates remain extremely low for the next two years. During this period, the growth in household borrowing rises from about 9% to 12%, so the Fed easing did facilitate borrowing.
September 2007-October 2008:
The Fed undid more than two years of rate increases in half as much time. This time it’s done little to improve access to credit, and may have actually reduced liquidity, as I argue above.

Conclusions: Fed rate increases successfully reduced household borrowing 1980-2, 1988-89, 1994. In 2000, it had an unclear effect. Since 2004, the Fed has lost control over credit creation.

The final chart explains why this has happened: The bond market has taken over lending in the economy. Contrary to popular belief, the banks don't lend money. The bond market does.


The dotted line in this chart shows the rise of securities the Fed calls "corporate bonds," which include normal corporates, non-GSE securitizations and financial-sector bonds. It's important to emphasize that it excludes Fannie and Freddie bonds and Treasuries. Most of the debt growth has come from non-GSE securitized debt (ie, toxic assets) and bonds issued by banks to fund their own lending activities. This highlights again the importance of capital markets, rather than normal bank lending, in the recent credit bubble. While some issuance is improving in the corporate bond market and the FDIC is backing bank bonds, I fear these efforts are still secondary, when they should be the primary focus. See my previous blog entry for more.

1 comment:

valar said...

EXXON TECHNOLOGIES also offers some interesting online job programs for all people. Everybody is interested to earn online through part time home based business opportunity. Start here your part time online home based business. We have guided thousands of members to achieve success using the online jobs. Some members are presently earning $300 - $500 and more per day, using our program and assistance. Millions of people in India & other countries are working Online Jobs. Now you have a great chance to work.Online jobs, only at www.exxontechno.com