The downside of an illiquid market.
Liquidity in the mortgage-backed securities market has declined significantly in recent years. Read why it happened and why it may be here to stay.
The aftermath of the financial crisis has witnessed a tremendous curtailment of risk within the financial services industry, particularly on Wall Street. This is hardly surprising, given the forceful regulatory, policy and legislative focus on reducing risk in in recent years. [paragraph] Higher capital requirements, limits on leverage and curbs on proprietary trading have drastically reduced profitability, especially for trading and market-making businesses, forcing several broker-dealers to either pull back or pull out from these businesses. [paragraph] While increased regulation has certainly achieved the policy objective of reducing risk, one unintended consequence for the mortgage market is reduced liquidity of agency mortgage-backed securities (MBS). [paragraph] To understand what this reduced liquidity means for lending, it's important to understand the root cause of this change. In particular, it's important to recognize that, despite being the most commonly cited cause of declining liquidity, financial regulation is neither the only driver nor the most important one. Several factors have contributed to the drop, most consequential of which is the unprecedented ramp up in the Fed's ownership of agency MBS since 2009.
We discuss these causes in detail in this article as well as our conclusion that current liquidity levels are here to stay. We also discuss that while current liquidity levels are no cause for immediate alarm, a substantial deterioration could have serious negative consequences for both lenders and borrowers.
What is liquidity and how is it measured?
Agency mortgage-backed securities are issued by Fannie Mae, Freddie Mac and Ginnie Mae. The vast majority of these are traded in the to-be-announced (TBA) market, which functions like a futures market where investors commit to buy or sell MBS that meet certain broad criteria. The specific securities delivered to the buyer are "announced" later, just before the settlement date, rather than at the time of trade.
The agency-MBS market is huge, with approximately $5.8 trillion in securities outstanding as of fourth-quarter 2015. This market has always been very liquid, as participants have been able to trade large volumes of securities relatively easily and quickly. As a result, plenty of buyers and sellers can transact without incurring large transaction costs or facing too much price volatility.
Liquidity can be measured along the four following dimensions--the first two are quantifiable; the last two are observed as trends and are open to interpretation:
* Transaction volume. Higher transaction volume indicates higher liquidity. Transaction volume measures total trading activity over a period of time and is commonly represented by the average daily trading volume, which reflects dollar volume of MBS transacted on a given day.
[FIGURE 1 OMITTED]
* Transaction costs. Higher transaction costs can be a sign of lower liquidity. Transaction costs, also known as the bid-ask spread, are the difference between the price market-makers pay the seller of a security and the price at which they sell the security to another buyer. This spread compensates market-makers for the cost (risk of adverse price movements, capital cost, profit margins) of warehousing the security. When markets are liquid, dealers are less worried about finding buyers for warehoused securities and more willing to provide market-making services to their clients, creating less price volatility. But when volatility is high, dealers are more exposed to the risk of adverse price movement on warehoused assets because buyers may not be readily available. This reduces competition and increases transaction costs.
* Resilience is the ability of markets to self-correct temporary price dislocations or mitigate volatility spikes so prices can return to normal quickly. Resiliency requires an active and heterogeneous investor base with divergent investment views.
* Depth refers to the ease of executing large trades. Dealers generally find it easier to warehouse smaller trades because of lower risk exposure and because buyers are more readily available. In contrast, large trades take longer because dealers have to work harder to find a buyer with a large appetite.
Is liquidity really down?
The average daily trading volume for agency MBS has declined in recent years--generally an indicator of worsening liquidity--but the current level also is a correction from a 2000-2008 ramp-up (see Figure 1).Today's daily volume of $200 billion also closely mirrors the 2004 level, just before the euphoria began.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
While a return to pre-crisis levels doesn't necessarily mean that current trading volume is at the right level, it does raise the question of whether liquidity was unsustainably high during the bubble and whether we are simply reverting to more sustainable levels.
To answer that question, we look at daily trading volume as a percentage of MBS outstanding, also known as turnover. Figure 2 shows that turnover also has declined since the housing crisis, but once again, only to pre-crisis levels. More importantly, this decline was also preceded by a sustained runup from 2000 to 2008, lending more credibility to the theory that there was too much liquidity previously.
Research from staff members at the Federal Reserve Board shows that the bid-ask spread for 3.0 percent, 3.5 percent and 4.0 percent coupon Fannie Mae MBS remained largely stable--5 to 7 basis points--between 2011 and 2013 (see Figure 3). However, notable here are the regular intervals at which the bid-ask spread has widened. Though the Federal Reserve Board's analysis does not extend past the end of 2013, market participants have informed us that today's bid-ask spread is well within the 5-to-7-basis-points historical range, suggesting no major deterioration other than periodic widening.
To summarize, the trading volume of agency MBS has clearly retreated since the housing crisis, but only to pre-bubble levels, supporting the excessive liquidity hypothesis. Further, the Federal Reserve Board's study shows that transaction costs have remained relatively stable except for occasional volatility.
These findings raise two new questions. Why is volatility up and why are trading volumes down?
[FIGURE 4 OMITTED]
Drivers of higher volatility
The primary driver of higher volatility is a major shift in MBS ownership from a large number of active traders to a smaller number of buy-and-hold investors post-crisis. This ownership shift has reduced both investor heterogeneity and the prevalence of contrarian trades, which diminishes the ability of markets to self-correct quickly, causing more volatility.
The data on MBS ownership pattern, plotted in Figure 4, reveal four trends:
* Unprecedented increase in the Fed's ownership of MBS. In its effort to stabilize the economy, the Federal Reserve began purchasing agency MBS in 2009 under the quantitative easing program. As a result, the Fed now owns more than $1.7 trillion in agency MBS, or roughly one-third of the total $5.8 trillion outstanding. Prior to 2009, the Fed owned no agency MBS.
* Commercial bank MBS holdings are up 50 percent. Commercial banks held less than $1 trillion in agency MBS in 2009. Today they own about $1.6 trillion. A tight lending environment has left banks with ample deposits that they have been investing in agency MBS instead.
Taken together, the Fed and banks now own about $3.3 trillion in agency MBS, or roughly 57 percent of total outstanding, compared with just 26 percent in 2006. Because both these entities are buy-and-hold investors, the tradable float--which is the amount of MBS available for buying and selling--has declined sharply. This affects ease of transacting because market-makers now have to work longer and harder to match buyers with sellers.
* Loss of the government-sponsored enterprises (GSEs) as a market lubricant. Historically, when markets were turbulent and spreads were wide, the GSEs were able to provide a bid--funding their purchases at lower borrowing costs because of an implicit guarantee--in order to profit from temporary price distortions. This intervention helped absorb volatility and brought prices closer to normal, effectively performing a market lubrication role.
But the terms of their bailout--which require a 15 percent annual portfolio reduction--as well as the intense political scrutiny of the portfolios have greatly diminished Fannie's and Freddie's ability and the desire to lubricate the market. Total agency MBS held in the GSEs' portfolios has shrunk from $770 billion in 2006 to $240 billion today, or by more than $500 billion. The GSEs owned about 22 percent of agency MBS outstanding in 2006 compared with only 4 percent today.
* Broker-dealer pullback. Dealers perform the role of middlemen by matching MBS buyers and sellers. Their effectiveness at this depends partly on the capital cost of holding MBS. Because of higher capital requirements, dealers now are more selective about which assets to own, how much to own and how long to hold them, effectively diminishing their ability to warehouse MBS in excess of levels they don't consider profitable.
Further compounding the problem is the inherently low profitability of the market-making business, given the tight bid-ask spreads. Pre-crisis, dealers were able to amplify profitability through leverage, which has also come under increased regulatory scrutiny. This has limited dealer efforts to grow balance sheets and hampered their ability to make markets as freely as before. Another reason is reduced competition among dealers thanks to the failure of Lehman Brothers, the sales of Bear Stearns and Merrill Lynch to JP Morgan and Bank of America, as well as the exit of Nomura, HSBC and Royal Bank of Scotland from the U.S. agency-MBS market.
[FIGURE 5 OMITTED]
According to New York Fed data on primary dealer market share, the top five dealers conducted about 55 percent of agency-MBS transactions in 2006. Today's top five account for about 80 percent. Reduced competition allows dealers to be pickier about which risks to take and which opportunities to pursue without fear of losing market share.
All together, dealers' and the GSEs' combined ownership of agency MBS has fallen from 27 percent of total outstanding in 2008 to less than 6 percent today. Because of the role these entities played in absorbing volatility, their pullback is felt the most during precisely such periods. The result is increased difficulty in executing large trades, more pronounced volatility and a longer duration before prices return to normal.
Drivers of falling trading volume
Trading volume tends to be positively correlated to new MBS issuance, which in turn is a function of mortgage originations. Strong issuance fuels more trading activity because a new security is typically bought and sold several times upon issuance, as short-term traders try to benefit from minor price imbalances before long-term investors step in and trading activity falls off.
Mortgage origination volumes have struggled in recent years for two reasons:
* The refinance boom is over. The vast majority of borrowers who were eligible and willing to refinance have already done so. Refinance originations comprised more than 80 percent of Fannie's and Freddie's and half of all Ginnie Mae's issuances from 2009 to 2013. But an ever-shrinking pool of eligible, in-the-money borrowers has caused refinance originations to decline since 2013 (see Figure 5).
* Struggling purchase originations. The void left by falling refinance volumes unfortunately has not yet been filled by a meaningful increase in purchase volumes. Extraordinarily tight credit standards have prevented millions of responsible borrowers from getting a mortgage in recent years. In addition, geographical mobility for both homeowners and renters has been declining for the past 25 years (see Figure 6).
What happens to liquidity next?
Even as capital markets have derisked and stabilized during the last few years, the aftereffects of the crisis continue to drive the volume and volatility trends described earlier. There are no concrete signs that these trends will reverse materially in the foreseeable future, leading us to believe that present levels of liquidity are here to stay.
First, the Fed's ownership share of outstanding agency MBS is unlikely to budge until the Fed changes course. Even though the quantitative easing program has ended, the Fed continues to reinvest principal paydowns. This means the Fed's holdings will remain constant at $1.8 trillion as long as this policy stays in place.
[FIGURE 6 OMITTED]
Second, the GSEs' bailout agreements require that the retained portfolios be downsized by 15 percent annually until they each hit $250 billion. This means Fannie Mae's current portfolio of $345 billion (which includes agency MBS, non-agency MBS and mortgage loans) needs to shrink an additional $95 billion, while Freddie's $350 billion portfolio needs to shrink another $100 billion. This will further diminish the role of two traditionally active market participants.
Lastly, there is no reason to expect any meaningful softening of the regulatory environment. Over the past several years, regulators have focused almost exclusively on derisking and there are no signs of a reversal. Perhaps more important, the last thing regulators want is to be accused of going easy on Wall Street.
Why should liquidity matter to lenders?
The entire foundation of the agency-MBS market is built on the confidence MBS sellers have in their ability to deliver securities on the settlement date--but more importantly, on buyers' uncompromising belief that sellers will make good on their commitment. And liquidity is the core mechanism that creates this trust.
As explained earlier, we expect current liquidity levels to persist for the foreseeable future, which is good news for lenders and borrowers. But a significant decline in liquidity from current levels that makes trading difficult or expensive could have serious negative consequences for borrowers, lenders and the mortgage market in general.
Lenders specifically can be affected in four ways:
* Higher costs and lower profitability. Any deterioration in the ease of trading will surely widen the bid-ask spread and immediately raise the cost of buying, selling or hedging MBS. This would ultimately affect profitability. Lenders, of course, can pass this cost on to borrowers (through higher rates) but competitive pressures could limit the extent. Higher rates could also price out some borrowers from the market and lead to lost originations.
* Greater interest-rate risk. Liquidity makes it possible for originators to short-sell MBS to hedge the risk that interest rates will fluctuate between when the mortgage application is received and when the mortgage is sold. This ability to hedge risk also allows borrowers to lock in their rates prior to closing. A worrisome drop in liquidity would make both of these benefits more difficult and more expensive to obtain.
* Higher price uncertainty. When markets are liquid, any new developments and events--both favorable and unfavorable--are priced in almost instantaneously, creating smoother price movements and lower volatility. A market that is less liquid is slower to price in new information, causing greater volatility when that information is suddenly accounted for.
* Less demand for agency MBS. The agency-MBS market attracts abundant capital from investors across the world. Global investors are hungry for agency MBS because these securities can be sold and converted to cash quickly whenever needed. If liquidity were to deteriorate substantially, many of these investors would likely switch to alternatives (such as Treasury securities). This would reduce the demand for agency MBS, causing securities to lose value and rates to rise.
The euphoria in the run-up to the financial crisis, which was caused by ever-increasing house prices, investor complacency, inadequate oversight and unchecked leverage, had artificially increased the demand for all asset classes, including agency MBS. The result was not only an asset price bubble, but also a liquidity bubble that has since burst. Despite the drop, however, liquidity remains generally healthy, albeit with occasional volatility caused primarily by more concentrated ownership of MBS--especially by the Fed--and to a lesser extent, by greater regulation.
Karan Kaul is research associate and Laurie Goodman is director of the Housing Finance Policy Center at the Urban Institute in Washington, D.C. They can be reached at email@example.com and firstname.lastname@example.org, and on Twitter[R] at @FollowKaranK and @MortgageLaurie. NOTE: To sign up for the Housing Finance Policy Center's bimonthly newsletter, email email@example.com.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||COVER REPORT: SECONDARY MARKET|
|Comment:||The downside of an illiquid market.(COVER REPORT: SECONDARY MARKET)|
|Author:||Kaul, Karan; Goodman, Laurie|
|Date:||May 1, 2016|
|Previous Article:||An untapped market.|
|Next Article:||Q&A with Congressman Ed Royce.|