Interest rate risk hurts — hurts like a 9 per cent market value loss | FT Alphaville

Interest rate risk hurts — hurts like a 9 per cent market value loss

The mortgage market — lurching from one risk to another, right?

No sooner had Fitch Ratings gotten more comfortable with credit losses than it starts warning on interest rate risk. It’s kind of back to the future for the Mortgage-Backed Securities (MBS) industry too. Because before the financial crisis, rate shifts were really the things keeping investors up at night.

Here’s Fitch’s macro credit team with a worked example in their latest report:

To take a practical example of the potential risks to MBS investors from higher interest rates, an unhedged Fannie Mae MBS (collateralized by 30-year fixed rate conventional mortgage loans) with a 3.5% pass-through rate incurred an approximately 9% price loss over the course of a few months, as Treasury yields increased 130 basis points (bp) between mid-October 2010 and mid-February 2011. By comparison, traditional, high-quality prime mortgages originated between 2000 and 2004 are expected to experience roughly 0.25% in cumulative credit losses, realized over the remaining life of the mortgage pools (e.g. up to 30-year term). Indeed, this 9% market value loss is nearly twice the cumulative expected credit loss of 5.0% for traditional prime mortgages originated between 2005 and 2008, the worst credit performance on record. Further exacerbating the risks to MBS investors in a rising rate environment is the potential for slowing prepayment rates, with many borrowers either unable (because of negative equity and/or tighter underwriting standards) or unwilling (because of escalating mortgage rates) to refinance.

There’s a rather nasty feedback loop that could happen here too, with MBS investors asking for higher yields to compensate for increased interest rate risk ending up increasing mortgage rates. There’s also a volatility issue which could lead to larger haircuts for repos collateralised by MBS. The MBS market is meant to be one of the most liquid US bond markets after US Treasuries. It’s a huge one too.

According to Fitch it’s roughly $6,800bn in size — with the biggest investors being US commercial banks with $1,260bn. Of course, the Federal Reserve has been fighting for first place thanks to its QEasing.

And the Fed angle is really where we’ve come across interest rate risk in MBS before, especially when it comes to prepayment of mortgages. To break it down, homeowners become increasingly reluctant to refinance their mortgages when interest rates start rising. So for MBS investors — like the Fed — it could take longer to get back their principal than they might initially have expected and accounted for.

And that’s where the potential problem for the Fed starts. Its whole borrow-short (0.25 per cent) lend-long (4 per cent) strategy could be in tatters, possibly harming its cashflow, once rates go up.

However — they have to go up enough.

Econbrowser recently summarised an economic letter from Glenn Rudebusch at the Federal Reserve Bank of San Francisco, saying the rate the Fed pays on reserves would have to rise to 7 per cent in order for the US central bank to start losing money on its current asset portfolio. Right now, that rate is just that quarter of a per cent. So Rudebusch thinks the Fed shouldn’t alter its easy monetary policy just because it’s worried about interest rate risk — that should be a very distant consideration.

Still, there are plenty of people out there who do worry about the risk impinging on the Fed’s independence. In fact, Rudebusch summarises some concerns right before his conclusion:

… Still, while it is generally recognized that central bank capital losses would not directly impede monetary policy operations, some analysts worry about the attendant political pitfalls (Borio and Disyatat 2010). Large realized or unrealized capital losses could be misinterpreted and subject the Fed to criticism, especially if the losses exceeded the additional interest income from the enlarged portfolio. In the worst case, the political backlash could perhaps threaten the Fed’s operational autonomy. In the past, the Bank of Japan has taken such threats quite seriously and has limited its balance sheet policy actions in part because of a fear that capital losses could tarnish its credibility (Bernanke 2003). Indeed, to insulate itself from such political fallout, the Bank of England obtained in advance an explicit government indemnity for potential future capital losses stemming from its program of large-scale asset purchases. The Fed’s accounting arrangements … automatically provide a similar implicit indemnification …

Luckily (ahem) the Fed has also already made it impossible for it to actually post capital losses.

Related links:
How big the BoE’s interest rate bind? – FT Alphaville
Is negative convexity the new Bernanke condundrum? – FT Alphaville
The Fed’s asset-liability mismatch – FT Alphaville, 2009