This page has been archived and commenting is disabled.
Fed Portfolio Duration Risk: $1.3 Trillion And Growing
The topic of the Fed's balance sheet has (rightfully) attained prominent status in recent weeks, due to the T-minus 3 months and counting until the last MBS are purchased on behalf of US taxpayers. Yesterday, the Federal Reserve issued an advisory on interest rate risk management that had the following cautionary language: "In the current environment of historically low short-term interest rates, it is important for institutions to have robust processes for measuring and, where necessary, mitigating their exposure to potential increases in interest rates." Ironically, it is none other than the Fed that due to its $1.8 trillion in outright security holdings may be best advised to heed its own warning, as it is on the hook for at least $1.3 trillion in interest rate risk.
As can be seen on the below chart, the Fed, in its quest to become the repository of every unwanted asset in existence, has accumulated over $1.8 trillion in assorted trash, with the bulk of maturities occuring in 2040, courtesy of its holdings of $908 billion in Agency MBS. Keep in mind these numbers will likely continue growing well past the $1.7 associated with just QE. We fully expect another $1 trillion in MBS to be purchased in QE 2.0 implying the Fed's balance sheet will likely come to rest at about $3 trillion in total security holdings (it also means you can throw out any hope for an efficient market out of the window for many years: if today's "good" NFP and consumer credit news are sufficient to result in an green close, we once again strongly urge you to pull all your money out).
Going back to the interest rate risk advisory, we note the following prophetic language:
Material weaknesses in risk management processes or high levels of IRR
exposure relative to capital will require corrective action. Such actions could include
recommendations or directives to:
- Raise additional capital;
- Reduce levels of IRR exposure;
- Strengthen IRR management expertise;
- Improve IRR management information and measurement systems; or
- Take other measures or some combination of actions, depending on the facts and circumstances of the individual institution.
We hope the Fed takes its own advice to heart. A simple analysis indicates that the duration of the Fed $1.8 trillion portfolio is over 17 years. In the warning, the Fed has said a move of 3-4% is to be modelled and expected:
As a result, institutions should regularly assess IRR exposures beyond typical industry conventions, including changes in rates of greater magnitude (e.g., up and down 300 and 400 basis points) across different tenors to reflect changing slopes and twists of the yield curve.
Applying a 4% IR shock to the Fed's portfolio results in 17 years x 4% x $1.8 Trillion = $1.27 Trillion. Furthermore, if as we expect the Fed's "assets" grow to $3 trillion, the portfolio risk becomes a stunning $2 trillion. Of course, the Fed can merely print money to bail itself out when it needs to.
Yet the Fed should certainly go through the abovementioned checklist of "recommendations and directives" and announce to US taxpayers, with whose money it gambles and continues to inflate the stock market to simply incredible heights, just what remedial actions it itself is taking to moderate Interest Rate Risk exposure.
h/t Mike C.
»
- 6379 reads
- Printer-friendly version
- Send to friend
- advertisements -
Well its OK 2040 comes after Y2K38.
Translation: The fed does not have the ability to effect any change in the money supply other than print. There are no means available to sop up excess dollars without making a complete mess of interest rates.
Welcome to the cash for trash program.
"There are no means available tos op up excess dollars without making a complete mess of interest rates."
Higher oil prices.
Would that explain what's going on in Yemen and Somalia? A strike against Iran would close the Straight of Hormuz, Saudi Oil would have to go East to West to Red Sea, right? Past Somalia and Yemen.
So how exactly would Higher Oil price help inflation? Where would that money go? China would have dollar up for oil? Someone please help me understand this, I've heard bits and pieces but not the whole theory.
Not to worry Tyler. The Fed will just hold the MBS until the collateral reaches maturity.
Don't you know yet that Agency Debt now has the "full faith and credit" of the USA for principal and interest? What another 39.7 years, anyway?
Good point. Now that the Fed pays interest on bank reserves, they will funding the assets to death at 4% + +.
The Agency MBS portfolio could have a long duration or after default zero.
Here's my plan...
I think I'm going to start a bank and securitize all the holdings in my storage unit - old furniture, busted toys, infinite paperwork, wife's decorations, rat turds, etc. - sell the JBS (junk backed securities) to the Fed for $1000/$1 valuation (gee, that's about right, isn't it???) and invest it all in the stock of the company that owns the storage unit.
Sometime later, I will write a book and give interviews on cnbc from the Azores.
Ahhh... the good life.
$1.3 trillion. Is anyone else starting to get numb to these huge numbers?
Simple - they manage IRR by crashing the stock market - Been discussed here several times OR they go in and load up on more of the same crap with YOUR money !!
HEY!!!! Isn't that right around the time SS & Medicare are bust???!!! China anyone?
I just love it when the Fed tells banks to
a)crank up the lending while
b)telling them to increase equity ratios and now increase liquidity for coming higher rate environment -- which is done in part by making fewer loans.
We are governed by morons. I do not see myself as a rocket scientist, but over the past 10 years I have increasingly believed they are representative of the majority of Americans. Sadly we may have the government we deserve.
The good news is that we'll never really "see" those loses as they won't ever open up their books. Likewise, if they aren't making the banks MTM their holdings, why in the hell would they do so?
Nice piece Tyler.
A simple analysis indicates that the duration of the Fed $1.8 trillion portfolio is over 17 years.
Tyler, you really have no idea of what you're talking about.
How on earth do you perform this "simple" calculation ?!?
I don't know what he's talking about here. Under current speeds I'd assume the duration to be around 5. He is correct in duration risk but I'm puzzled on that "simple" and "17 year" duration figure.
Applying a 4% IR shock to the Fed's portfolio results in 17 years x 4% x $1.8 Trillion = $1.27 Trillion.
Bravo Sharts! Woe to he who presumes a 400 Bp spike at 20 years on the curve. Dig out the ol' Sharpe textbook TD.
Tyler, the duration of the Fed's portfolio is nowhere near 17 years. It looks like you are using the legal final maturity for their agency MBS. Mortgages prepay (remember that whole negative convexity thing?). They also amortize. The actual duration of their MBS assets is probably closer to 10 years, not 30.
The actual duration of their MBS assets is probably closer to 10 years, not 30.
With a mix of 15 and 30 year MBS and a lot of coupons that are above the current mortgage rates (though refinancing is tougher since a lot of homeowners don't have enough equity) I'd think the duration is a lot closer to 5 years than 10. Tyler is only about 3-4X too high with his "simple calculation".
Fair enough. A lot of it depends on your prepayment assumptions. I was just pointing out that using the legal final maturity makes zero sense and that the duration of the Fed's MBS is much shorter than 30 years.
That's assuming the stuff they bought is worth more than 0.
Like above - it is awfully misleading to show the majority of maturities in 2039 as very few borrowers will still have an outstanding mortgage at that point. This post is scary as basic fixed income knowledge seems to be missing. Somewhere Harley Bassman is smiling
And the tiny percentage of 30 year mortgages that do have any balance left in 2039 will be making their last few payments and have a tiny sliver of original face value left.
"This post is scary as basic fixed income knowledge seems to be missing."
Really??
Tyler is estimating the maximum rate risk - not valuing the mortgages!!
When mortgage payments are made and refinancing proceeds are paid under the Fed's MBS holdings, the purchase of T-bills on the open market with those proceeds will drain reserves, thus tightening! To avoid draining reserves the Fed is going to have to reinvest principal and interest - most probably in new MBS to prop up housing prices and to hold longer term rates in check.
Given the Fed's policy goals and present actions one could reasonably assume that rate risk of similar duration will be assumed with each roll into yet new MBS and that the "duration" of its mortgage related assets is infinite! But 17 years is as good a guess as any. :)
Welcome to the Mad Hatter's tea party!!
Please pardon me for two probably ridiculous questions: can the Fed just simply continue to soak up bad assessts and trash and hold on to it indefinately, on its balance sheet, without consequence? When does the Fed become insolvent?
The duration is much shorter than 17 yrs. The FED's mtg portfolio has a worst case effective duration of 5 yrs in the base case and due to its negative convexity - perhaps around 6.5% if rates back up.
A little off subject here, but are mortgage voluntary pre-payments slowing down or hanging in?
Not an MBS expert.
they are very low due to home price depreciation. Most pre 2009 origination have current ltv over 80 and 40% of 2006/07 may be 100 or higher. Even though gse have the credit risk , these borrowers are having a very difficult time refinancing. The HARP program along with FED/TREAS buying was suppose to help but has been a complete failure. So the FED/TREAS bought 1.5 trillion mtgs in an effort to help purchases and refi's (stimulus and reduces credit risk for GSE by lowering borrowers monthly payment without increasing debt but they forgot to tell the banks/gse's to reduce their fees to make it a success. Bottom line, a giant wealth transfer from taxpayers to MBS holders (lack of refi makes premium mbs more valuable).
So the paper is impaired, most likely horribly so. But how bad will underlying defaults be? 20% seems incredible, but the paper will still have performing assets that will generate some yield.
How likely is it that sellers to the Fed have cherry picked the underlying to give the Fed with only underlying with negative equity and/or delingency issues?
from a credit perspective, the credit of the underlying fn/fh MBS makes no difference to the investor. The GSE (now the the US Treasury) are obligated to pay the MBS holders timely payment of principal and interest if the underlying mtgs are delinquent or default. When a default occurs, the GSE typically buys the loan from the MBS pool at par which in effect is a prepayment to the MBS holder. The delinquent loan then ends up on the GSE BOOKS at some value less than par - say 50c and the GSE takes the loss. In theory, the FED only has interest rate risk - no credit risk so cherry picking the loans is not relevant.
To answer the questions regarding ultimate default rates at the GSE, probably will end up around 10-12% of all loans guaranteed during the 2004-2008 period - assuming home prices stabilize at current levels. Losses could go as high as 400bb. The good news is that the loans from 2009 have very good credit characteristics -low ltv/high fico. These loans will travel with the GSE when they are spun off in the good/bad bank solution. Taxpayers will be left with the pre 2009 garbaGE. The bad news is that the FHA is underwriting most of the new marginal loans -in VERY LARGE SIZE - and will really get creamed if prices fall 10% from here. These 97% ltv loans have NO cushion.
Now, all of the above does not take into account the growing risk/popularity of strategic defaults -people who can pay but just walk away. If this gain momentum, it is lights out for FHA and the banking sector. FHA does not model this risk and when u lend at 97 ltv, you are very exposed to a small decline in housing values.
Hope this was helpful.
Thx. Very helpful.
So if the Federal Reserve has bought all these MBS, that means that they own all the real estate that that they are based on. Sounds to me like the Federal Reserve is going to have to become a major employer. They're going to need thousands of people to maintain, prepare and sell all that property.
I don't see people queing up to buy housing in any numbers in the near future so they are going to have to maintain them for a long time.
So. Public Utilities, Property Taxes, Insurance all paid by the federal reserve.
Mind boggling.
Tyler, your duration assumption makes absolutely no sense. What's the mix of floating vs. fixed paper etc?
I have to say, ZH standard has deteriorate dramatically over the past six months or so. Please apply the same standard to your analysis that you expect from others (e.g. the Fed, Treasury, MSM etc).
I wonder about the facts in this post myself. I saw references to this advisory published in 2 different places last week and as I remember it was both times ascribed to FDIC not the Fed. I can't find the references now. Am I wrong?
If the value of the MBSs fall, then a chunk has to be taken out of the assets in the Fed's balance sheet. [H.4.1]. That fall in assets has to be balanced by a equal fall in liabilities somewhere, else the books are out of balance.
Had the Fed been a regular corporation, the chunk would have come out of shareholders' equity - but the Fed isn't.
So what would happen? Would reserve balanced be lowered? Or, is there shareholders' equity in the Fed's full balance sheet which would be knocked down? If the latter, would the Fed stop returning its above-dividend profits to the Treasury for a time? Will it stop paying dividends too?
These questions, I think, are hypothetical because the MBSs are guaranteed by Ginnie Mae, Freddie Mac and Fannie Mae. So, what would happen if the MBSs became impaired is those agencies would make the Fed whole.
But the questions would become real if the Fed were ever forced to mark their assets to market. The books are put together on the assumption that the assets the Fed holds will never decline in value. If declines were allowed, then we're in completely uncharted territory.
Conclusion: the Fed will never have to mark to market. Any attempt to make the Fed do so will be fought tooth and nail, and will likely be defeated.
Looks like TPTB have kept their word. There is now a "bad bank". It is the Bank of YouAndMe. All the junk will be loaded off to Fannie, Ginnie, and Freddie for us to resolve. Ain't finance wonderful in the miracles it performs?
DM Ryan - MTM will happen by itself as housing sinks an additional 25%....