My last piece on this topic, On Bond Market Illiquidity (and more), drew a few good comments. I would like to feature them and answer them. Here’s the first one:
One issue you don’t address in your post, which is excellent as usual, is the impact of what I’ll call “vaulted” high quality bonds. The explosion and manufacturing of fixed income derivatives has continued to explode while the menu of collateral has been steady or declining. A lot of paper is locked down for collateral reasons.
That’s a good point. When I was a bond manager, I often had to deal with bonds that were salted away in the vaults of insurance companies, which tend to be long-term holders of long-term bonds, as they should be. They need them in order to properly fund the promises that they make, while minimizing cash flow risk.
Also, as you mention, some bonds can’t be sold for collateral reasons. That can happen due to reinsurance treaties, collateralized debt obligations, accounting reasons (marked “held to maturity”), and some other reasons.
But if the bonds are technically available for sale, it takes a certain talent to get an insurance company to sell some of those bonds without offering a steamy price. You can’t sound anxious, rushed, etc. My approach was, “I’d be interested in buying a million or two of XYZ (mention coupon rate and maturity) bonds in the right price context. No hurry, just get back to me with any interest.” I would entrust this to one mid-tier broker familiar with the deal, who had previously had some skill in prying bonds out of the accounts of long-term holders before. I might have two or three brokers doing this at a time, but all working on separate issues. No overlap allowed, or it looks like there is a lot of demand for what is likely a sleepy security. No sense in driving up the price.
Because it is difficult to get the actual cash bonds, it is tempting for some managers to buy synthetic versions of those bonds, or synthetic collateralized debt obligations of them instead. Aside from counterparty risk, the derivatives exist as “side bets” in the credit of the underlying securities, and don’t provide any additional liquidity to the market.
My point here would be that these conditions have existed before, and I think what we have here is a repeat of bull market conditions in bond credit. This isn’t that unusual, and it will eventually change when the bull market ends.
Here’s the next comment:
I hope you’re doing well. I’ve been reading your blog for about a year now and really appreciate your perspective and original content. Just wanted to ask a quick question regarding your most recent post on bond market liquidity.
Our investment committee often talk about the idea of bond liquidity (and discusses it with every bond manager who walks in our doors), and specifically how there are systematic issues now which limit liquidity and considerably push the burden onto money managers to make markets vs. the past, when banks themselves were free to make more of a market with their own balance sheets.
My (limited) understanding is that legislation since 2008 has changed the way that investment banks are permitted to trade on their own books, and this is a big part of the significantly decreased liquidity which has thus far been a relative non-issue but which could rear its head quickly in the face of a sharp correction in bonds.
Do you have any thoughts about this newer paradigm of limited market-making at the big banks? You didn’t seem to mention it at all in your article and I’m wondering if my thoughts here are either inaccurate or not impactful to the bottom line of the liquidity conversation.
I’m sure you’re a busy guy so I won’t presume upon a direct response but it may be worthwhile to post an update if you think these questions are pertinent.
Another very good comment. I thought about adding this to the first piece, but in my experience, the large investment banks only kept some of the highest liquidity corporates in inventory, and the dregs of mortgage- and asset-backed bonds that they could not otherwise sell. The smaller investment banks would keep little-to-no-inventory. Many salesmen might have liked the flexibility of their bank to hold positions overnight, or buying bonds to “reposition” them, but the experiences of their risk control desks put the kibosh on that.
As a result, I think that the willingness of investment banks to make a market rely on:
- The natural liquidity of the securities (which comes from the size of the issue, market knowledge of the issue, and composition of the ownership base), and
- How much capital the investment bank has to put against the position.
The second is a much smaller factor. Insurance companies have to deal with variations in capital charges in the bonds that they hold, and that is not a decisive factor in whether they hold a bond or not. It is a factor in who will hold a bond and what yield spread the bond will trade at. Bonds tend to gravitate to the holders that:
- Like the issuer
- Like the cash flow profile
- Have low costs for holding the bonds
Yes, the changed laws and regulations have raised the costs for investment banks to hold bonds in inventory. They are not a preferred habitat for most bonds. Therefore, if an investment bank buys a bond in order to sell it (or vice-versa) in the present environment, the bid-ask spread must be wider to compensate for the incremental costs, thus reducing liquidity.
To close this evening, one more letter on bonds from a reader:
First off, thank you for taking the time to share your knowledge via your blog. It is much appreciated.
Now for a bond question from someone learning the fixed income ropes…
What is the advantage/reasoning behind a company co-issuing notes with a finance subsidiary? Even with reading the prospectus/indenture I can’t understand why a finance sub (essentially just set up to be a co-issuer of debt) would be necessary especially since the company is an issuer anyway and they also may have other subs guarantee the debt also. I’m probably missing something obvious.
The answer here can vary. Some companies guarantee their finance subsidiaries, and some don’t. Those that don’t are willing to pay more to borrow, while bondholders live with the risk that in a crisis, the company might step away from its lending subsidiary. They would never let the subsidiary fail, right?
Well, that depends on how easy it is to get financing alternatives, and how easy it might be for the parent company to borrow, post-subsidiary default.
If things go well, perhaps the subsidiary could be spun off as a separate company, or sold to another finance company for a gain. After all, it has had separate accounting done for a number of years.
Beyond that, it can be useful to manage lending separately from sales. They are different businesses, and require different skills. Granted, it could be done as two divisions in the same company, but doing it in separate companies would force separate accountability if done right.
There may be other reasons, but they aren’t coming to my mind right now. If you think of one, please note it in the comments.