James Hamilton of Econbrowser writes another marvellous review of the Fed’s Balance sheet, updating his prior commentary which was also reviewed on PrefBlog.
One thing I had been unclear about was the precise nature of the “Supplementary Financing Program Account” of Treasury at the Fed, which is financing all the special programmes. While the assertion has been made that the Fed’s intervention is sterilized (meaning that it is causing no increase in monetary aggregates) … I wasn’t sure. However, the Monthly Treasury Statement referenced by Dr. Hamilton shows clearly (Table 6 on page 20) that Treasury has issued about $630-odd billion in Treasury Securities in fiscal 2009 to date, of which $588-billion has been from the public. This more than covers the $134-billion increase in the Supplementary Account, while still leaving $402-billion to finance the deficit … which is the total deficit for F2009 reported in Table 5 on page 18.
OK, so that’s cleared up!
Dr. Hamilton concludes:
For the record, let me reiterate my personal position on all this.
(1) I am doubtful of the Fed’s ability to alter interest rate spreads through the kinds of compositional changes in its balance sheet implemented over the last two years. Whatever your prior ideas were about this, surely it’s time to revise those in light of incoming data– if the first trillion dollars didn’t do the job, how much do you think it would take to accomplish the task?
(2) I think the Fed’s goal should be a 3% inflation rate. Paying interest on reserves and encouraging banks to hoard them is inconsistent with that objective, as would be a new trillion dollars in money creation.
I would therefore urge the Fed to eliminate the payment of interest on reserves and begin the process of replacing the exotic colors in the first graph above with holdings such as inflation-indexed Treasury securities and the short-term government debt of our major trading partners.
I’m not sure that point 1 is phrased in a useful manner. As I see it, the objective is not so much to maintain spreads as it is to ensure that the market exists at all. It has been observed that securitization has declined, which has had essentially forced banks to intermediate between borrowers and lenders, as opposed to simply engaging in the disintermediation inherent in packaging their loans and taking the spreads and servicing fees.
John Kiff, Paul Mills & Carolyne Spackman wrote a piece for VoxEU, European securitisation and the possible revival of financial innovation:
Collapsing global securitisation volumes in the wake of the subprime crisis have raised fundamental questions over the viability of the originate-to-distribute business model.1 Issuance has dropped precipitously in both Europe and the US, with banks keeping more loans on their balance sheets and tightening lending standards as a result (Figure 1). The decline has been particularly sharp for mortgage-backed securities and mortgage-backed-securities-backed collateralised debt obligations. The originate-to-distribute model was thought to have made the financial system more resilient by dispersing credit risk to a broad range of investors. Ironically, however, it became the source of financial instability.
…
The risk transfer and capital saving benefits of securitisation, combined with underlying investor demand for securities, should eventually revive issuance. But the products are likely to be simpler, more transparent, and trade at significantly wider spreads.
All that lost securitization issuance is staying on banks’ balance sheets and there are only a few possibilities:
- let the market collapse: in this case, banks will simply cease to make new loans; their balance sheets won’t take the strain and they can’t really issue new equity while the markets are so awfully depressed without really sticking it to their existing shareholders, or
- let the markets adjust.
An adjustment in the market can take place in several different ways:
- banks can re-intermediate: this will require balance sheet expansion, which can’t happen until they can sell equity at reasonable prices, or
- securitization markets can get restarted
I suggest that it is a Public Good for securitization markets to restart and agree with Kiff et al. that this will likely be accompanied by greater transparency and wider spreads. Trouble is, nobody knows what those spreads will be like.
What should the spread of mortgages over governments be? Agency spreads in the US were minimal prior to the current crisis and it seems clear to me that they should be wider. I’ve tried to find an easy graph for mortgage spreads in Canada – where securitization is nowhere near as important – but the best I’ve been able to come up with is a chart from 1999:
Mortgages are not, perhaps, the best example to choose because as I have repeatedly noted, there is a lot more that’s wrong with the American mortgage market than mere sub-prime:
Americans should also be taking a hard look at the ultimate consumer friendliness of their financial expectations. They take as a matter of course mortgages that are:
- 30 years in term
- refinancable at little or no charge (usually; this may apply only to GSE mortgages; I don’t know all the rules)
-
non-recourse to borrower (there may be exceptions in some states)
- guaranteed by institutions that simply could not operate as a private enterprise without considerably more financing
- Added 2008-3-8: How could I forget? Tax Deductible
Clearly, in the particular case of US mortgages, the underlying pools must not just trade at wider spread, but they must be more investor friendly.
However, I do recognize Dr. Hamilton’s desire to put a limit on the amount of reintermediation that is being done by the Fed, but must disagree with the prescription of keeping the balance sheet grossed up with government bonds of any description.
I suggest that a schedule be put into place whereby, for instance, the Commercial Paper Funding Facility have its spreads gradually widened. Rates are now 2.19% for unsecured commercial paper and is scheduled to cease purchasing new paper on April 30, 2009. The current rate paid on excess & required reserve balances is now 0.25%. Thus, it is apparent that in the current environment, a spread of 194bp is not enough to get the banks to move into commercial paper in a big way. The same applies to the general public.
I suggest that this is a distress-level spread, being paid for CP of perfectly good quality; indicating a flight to safety. Eventually the climate of blind fear will dissapate, but until that happens the Fed should continue to apply the implicit Bagehot prescription of making credit freely available at punitive rates. And, perhaps, announce that the programme will be extended past April, but at spreads on CP of 110+110 (for the duration of the extension), rather than the current 100+100. Eventually, one of several things will happen:
- Greed will overcome fear, and banks (et al.) will cease lending to the Fed at 0.25% and start lending to solid companies at 2.25%, or
- Companies will refinance with longer term paper, or
- Companies will go bust.
What is the YTW of RY.PR.N? Win a PrefLetter!
December 22nd, 2008I will admit that sometimes I look at the analysis generated by HIMIPref™ and blink. The assumptions and procedures and approximations used in the course of the analysis can sometimes work together in unexpected ways … so the results need to be reviewed in order to determine whether
Such are the joys of quantitative analysis, when you can spend a month trying to figure out the analysis of one instrument on a date from ten years back!
This time, however, it’s today’s analysis of RY.PR.N: it closed today at 26.00-10, 28×1, after trading 29,390 shares in a range of 26.00-10.
And yet despite the $26.00 price, HIMIPref™ shows the pre-tax bid-YTW scenario as being the limitMaturity – that is, the dummy maturity thirty-years hence which is used as a substite for “forever”.
First, some facts: the issue closed on December 9 and is a fixed reset with the terms 6.25%+350. The analysis assumes that 5-year Canadas will now and forever yield 1.83%, so the rate is presumed to be reset to 5.33% at the first (and all subsequent) reset dates.
HIMIPref™ calculates the yield to first call of 5.4130% and yield-to-limit of 5.2913%. I have uploaded the cash-flow reports for the five year and 30-year maturities. The YTW is the worst yield, 5.2913%, and the YTW scenario is the 30-year maturity.
There cannot be much argument about the yield calculation for the five year maturity; everything is known, so it’s all perfectly standard. However, the thirty year maturity is simply an analytical placeholder for “forever” and the maturity value is not known. As you can see from the reports, HIMIPref™ estimates a price of $23.44 for the 30-year case.
Why $23.44? For that we have to look at the HIMIPref™ calculation of costYield … I have uploaded the relevant cash flow analysis. Readers will note the cash flow entry dated 2014-3-26, for -1.73 (future value) discounted to -1.34 (present value). This is the estimate of what the issuer’s call option is costing the holder; the implication is that if this option didn’t exist, we’d be willing to pay $1.34 (present value) more for the security.
The value of the option is calculated using a time-influenced distribution of possible prices centred on the current price. As shown by the Option Cash Flow Effect Analysis, it is currently assumed that there is a 53% chance of the option being exercised. Slicing the price distribution into two parts on that date, it is calculated that the average unconstrained price in exercise scenarios is 28.24; the average unconstrained price in non-exercise scenarios is 23.44. Voila! An estimated maturity price of $23.44.
I’ve also uploaded an Excel spreadsheet where I did a little fooling around with the reports. Raw data is in cells a1:e128. I’ve converted the semi-annual yield back into annual in cells c129:c130. The cash-flows with some decimals put back in are in cells g1:g122. My check on the arithmetic is in cells i1:j122 and sum to a present value of $26.03805; I’m assuming that the extra 3.805 cents is due to rounding differences of dates and days-in-year approximations. I used cells l1:n124 to play around with the yield-effect of different maturity values, and summarized my playing in cells l127:n130, which I will reproduce here:
Effect of Maturity Value
on Calculated Yield
It’s not all that sensitive, but the rate with a 26.00 end-value is slightly in excess of the 5-year rate, implying that if we rely on a 26.00 end-value then the 5-year yield is the YTW … as would be expected.
But I claim that you cannot count on a 26.00 end-value. I claim that if the unconstrained market price is 26.00 on a call date, then the issuer will call the issue at 25.00 instead. All you can count on at the end of eternity (which is 30-years off) is that fraction of the price distribution that escaped the calls … and that has an average value of 23.44.
And hence, the YTW scenario for a 26.00 issue callable at 25.00 in five years is … the limit maturity. This doesn’t happen for normal “straight” perpetuals: if the issue had an expected cash flow stream of 6.25% for the entire 30-year period, rather than 6.25% for five years and 5.33% thereafter, the five-year call would have a lower yield and hence be the YTW scenario.
And, just for fun, let’s have a contest! Presuming an end-value of 23.44, what post-reset 5-Year Canada yield (and hence, what dividend rate on RY.PR.N) do we need to bump the yield up to the point where the 5-year call becomes the Yield-To-Worst scenario? First correct answer wins a copy of the January edition of PrefLetter.
Posted in Issue Comments, PrefLetter | No Comments »