The Education of a Mortgage Bond Manager, Part I

You might remember my “Education of a Corporate Bond Manager” 12-part series.  That was fun to write, and a labor of love, but before I was a corporate bond manager, I was a Mortgage Bond Manager.  There is one main similarity between the two series — I started out as a novice, with people willing to thrust a promising novice into the big time.  It was scary, fun, and allowed me to innovate, because in each case, I had to rebuild the wheel.  I did not have a mentor training me; I had to figure it out, and fast.  Also, in this era of my career, I had many other projects, because I was the investment risk manager for a rapidly growing life insurer.  (Should I do a series, “The Education of a Financial Risk Manager?”)

One thing my boss did that I imitated was keep notebooks of everything that I did; if this series grows, I will go down to the basement, find the notebooks, and mine them for ideas.  When you are thrust into a situation like this, it is like getting a sip from a firehose.  Anyway, I hope to do justice to my time as a mortgage bond manager; I have been a little more reluctant to write this, because things may have changed more since I was a manager.  With that, here we go!

Liquidity for a Moment

In any vanilla corporate bond deal, when it comes to market for its public offering, there is a period of information dissemination, followed by taking orders, followed by cutoff, followed by allocation, then the grey market, then the bonds are free to trade, then a flurry of trading, after which little trading occurs in the bonds.

Why is it this way?  Let me take each point:

  1. period of information dissemination — depending on how hot the market is, and deal complexity, this can vary from a several weeks to seven minutes.
  2. taking orders — you place your orders, and the syndicate desks scale back your orders on hot deals to reflect what you ordinarily buy and even then reduce it further when deals are massively oversubscribed.  When deals are barely subscribed, odd dynamics take place — you get your full order, and then you wonder, “Why am I the lucky one?”  After that, you panic.
  3. cutoff — it is exceedingly difficult to get an order in after the cutoff.  You have to have a really good reason, and a sterling reputation, and even that is likely not enough.
  4. allocation — I’ve gone through this mostly in point 2.
  5. grey market — you have received your allocation but formal trading has not begun with the manager running the books.  Other brokers may approach you with offers to buy.  Usually good to avoid this, because if they want to buy, it is probably a good deal.
  6. bonds are free to trade — the manager running the books announces his initial yield spreads for buying and selling the bonds.  If you really like the deal at those spreads and buy more, you can become a favorite of the syndicate, because it indicates real demand.  They might allocate more to you in the future.
  7. flurry of trading — many brokers will post bids and offers, and buying and selling will be active that day, and there might be some trades the next day, but…
  8. after which little trading occurs in the bonds — yeh, after that, few trades occur.  Why?

Corporate bonds are not like stocks; they tend to get salted away by institutions wanting income in order to pay off liabilities; they mature or default, but they are not often traded.

By this point, you are wondering, if the title is about mortgage bonds, why is he writing about corporate bonds?  The answer is: for contrast.

  1. period of information dissemination — depending on how hot the market is, and deal complexity, this can vary from a several weeks to a few days.  Sometimes the rating agencies provide “pre-sale” reports.  Collateral inside ABS, MBS & CMBS vary considerably, so aside from very vanilla deals, there is time for analysis.
  2. taking orders — you place your orders, and the syndicate desks scale back your orders on hot deals to reflect what you ordinarily buy and even then reduce it further when deals are massively oversubscribed.  When deals are barely subscribed, odd dynamics take place — you get your full order, and then you wonder, “Why am I the lucky one?”  After that, you panic.
  3. cutoff — it is exceedingly difficult to get an order in after the cutoff.  You have to have a really good reason, and a sterling reputation, and even that is likely not enough.
  4. allocation — I’ve gone through this mostly in point 2.
  5. grey market — there is almost no grey market.  There is a lot of work that goes into issuing a mortgage bond, so there will not be competing dealers looking to trade.
  6. bonds are free to trade — the manager running the books announces his initial yield spreads for buying and selling the bonds.  If you really like the deal at those spreads and buy more, you can become a favorite of the syndicate, because it indicates real demand.  They might allocate more to you in the future.
  7. no flurry of trading — aside from the large AAA/Aaa tranches very little will trade.  Those buying mezzanine and subordinated bonds are buy-and-hold investors.  Same for the junk tranches, should they be sold.  These are thin slices of the deal, and few will do the research necessary to try to pry bonds out of their hands at a later date.
  8. after which little trading occurs in the AAA bonds — yeh, after that, few trades occur.  Same reason as above as for why.  Institutions buy them to fund promises they have made.

Like corporate bonds, but more so, mortgage bonds do not trade much after their initial offering.  The deal is done, and there is liquidity for a moment, and little liquidity thereafter.

Again, if you’ve known me for a while, you know that I believe that liquidity can’t be created through securitization and derivatives.  Imagine yourself as an insurance company holding a bunch of commercial mortgage loans.  You could sell them into a trust and securitize them.  Well, guess what?  Only the AAA/Aaa tranches will trade rarely, and the rest will trade even more rarely.  The mortgages are illiquid because they are unique, with a lot of data.  You would have a hard time selling them individually.

Selling them as a group, you have a better chance.  But as you do so, investors ramp up their efforts, because the whole thing will be sold, and it justifies the analysts spending the time to do so.  But after it is sold, and months go by, few institutions have a concentrated interest to re-analyze deals on their own.

And so, with mortgage bond deals, even more than corporate bond deals, liquidity is but for a moment, and that affects everything that a mortgage bond manager does.  More in part 2.

 

7 Comments

  • Conscience of a Conservative says:

    With mortgages there are only so many servicers and originators and vintages. The guys buying and holding this paper know the routine. Once they understand who sold it and services it, it’s just a mater of re-examining how the loans in this particular deal are transitioning and then deciding on a price for that exposure should you want it.

  • Doug says:

    David, your claim that securitization adds no liquidity is reminiscent of Zeo’s Paradox of Achilles and the Tortoise. Supposedly, Achilles could never reach a tortoise in a footrace, because every time he ran to where the tortoise was, the tortoise would have moved.

    The problem is one of motion. In the paradox, Achilles slows continuously, whereas in real life, runners move forward at a constant rate. In financial securitization, just because a bond isn’t infinitely traded doesn’t mean there’s no improvement in the liquidity.

    First, define liquidity. Then measure whether it has improved. I think a fair measure of liquidity is the bid/ask spread. I don’t think anyone will argue that the bid/ask spread for MBS is wider than it is for unsecured whole-loans. Nor do I think the bid/ask spread for MBS collateral is wider than that of corporate bonds.

    It’s the fallacy of the excluded middle: either you have daily trading and tight markets, or you have no improvement in liquidity. Wrong. There are gradations of liquidity, just as there are to most things in life.

    • My experience is that if you create a passthrough of a bunch of loans that are not that liquid, if you tranche it, only the AAAs are slightly liquid, and all junior tranches are not liquid.

      My opinion is that form does not affect liquidity, only the liquidity of the underlying affects liquidity.

  • you mentioned looking for long ideas – consider FBIZ.

    http://snowleopardinvesting.wordpress.com/

    great blog.

    • Thanks. If you want, I can put you in touch with some friends who are experts on small banks. Neither of them own FBIZ. One is small enough to do so — small financials, whether insurers or banks, tend to get low valuations. Size is usually a positive valuation factor up to a point, then it is a negative.

  • Thanks for the offer but I’ll decline for now as I’m definitely not an expert. I agree about your valuation comment, but it appears that many similarly sized banks (with worse track records) are trading at 12x normalized earnings or more. I’m working on a post about the sector now. again, love the blog!