Why an exchange model won't work for corporate bonds

Here is some data that always surprises equity investors about the difference in the liquidity profiles between the equity and debt markets.

The US equity market has ~6k publicly traded listings of which ~99% trade daily.

The US corporate market has ~600k issues of which only 3% trade in a given week.  

Please note the obvious....an issuer typically only has 1 common equity....but could have dozens or even hundreds of senior unsecured debt issues.    

If you look at the top 25 most active corporate bond issues, they typically make up about 21% of the markets overall volume.  Typically, these issues are recent new issues.  It often seems to investors that the major underwriters are bringing new issues to market, trading them briefly in the secondary and then they go to 'bond heaven' (or hell as the case may be).

So, liquidity in the credit markets is very concentrated in a few recent new issues...and beyond these issues, the volume thins dramatically.  I've seen numerous instances where ~$5oomm deals don't trade even 1 bond for weeks and sometimes months on end.

As an aside, the sectors that habitually have the worst liquidity are REITs, insurance and bonds with a maturity of less than 5 years (all sectors).  

So, say a Rohm and Haas bond with a 2029 maturity fits the profile of a potential buyer that wants to asset/liability match....but the only seller in the market of that issuer is of the Rohm and Haas 2020 maturity bonds....nary the twain shall meet.  There would not be a trade.  This is what would occur in an 'exchange model'.  Please note, that I am using the term 'exchange model' but the same holds true for a matching network model as well.  

In the 'old days', when dealers devoted capital to trading corporate bonds, a dealer would buy the 2020s in the market and (short) sell the 2029s to the client that needed them.  The 9 maturity difference would not add too much risk to their balance sheet.  The shift from principle to agency trading has virtually eliminated this service.  I've said this once before, you cannot have liquidity without capital.

There are 2 hurdles to be overcome if an exchange model is to succeed.  

One is investor preference.  It's been my experience that investors often enter the market looking for a particular bond....and not looking for 'what's available at the right price'.  There's often a good reason why only particular bond (or risk profile) is needed.  A reasonable investor does not want to accumulate a mish-mosh of small positions in one issuer (which is what could occur under the 'what's available at right price').  Given the thin liquidity profile of the bond market that I described earlier, this existing investor preference would need to change.  

Two, issuers need to reduce the number of different bonds outstanding in their name.  Re-opening an existing issue may cause a less than optimal asset/liability match for them but it will enhance the liquidity in their name.  This will allow them better access to the capital markets in the future.  

There is little economic incentive for the major participants (issuers, investors and dealers) to adapt to the measures mentioned above.  To state the obvious, most folks have far greater and more imminent problems to focus on.  Until this dissipates, don't expect liquidity to improve or an exchange model to develop.