On April 28th I published a piece on the FHLBs. Later I was contacted by the FHFA (the regulator for the FHLBs). They pointed to a factual error in my post and asked me to publish a correction.
I said the following:
Fannie (FNM) and Freddie (FRE) will cost us at least $400 billion. Add in another $100b for the FHLBs.
It is not correct that the FHLBs will suffer total losses of $100b. The FHFA provided a link to a balance sheet of the FHLB’s. The following is a slide of that report. As you can see (click to enlarge) there is a line item marked investments of $284B.
I relied on a the following sentence from an FHFA report to evaluate the investment portfolio:
FHLBanks' investment portfolios consist primarily of MBS securities, chiefly Agency MBS or highly rated private-label MBS.
When I think of the term, “private label MBS” I immediately think losses. And it is correct that the FHLBs have and will continue to suffer losses from these holdings. But the amount of the PL MBS in portfolio was substantially less than I had assumed.
The clarification provided by FHFA on this is as follows:
As discussed, the key point is that private-label MBS are $46 billion (not the $284 billion you had in the blog), and that the FHLBanks have already recorded credit charges and non-credit other-than-temporary impairments on these securities.
Some thoughts on this:
- I try quite hard to stay on track with my writing and this was a pretty serious error. I used a number that was not properly vetted and this resulted in a bad conclusion.
The broader issue of blogs and the accuracy of what is in them has been raised on a number of occasions. I have no editor that is looking over my shoulder. Nor do many of the other bloggers who produce the 200-300 articles on finance that are appearing every day. Yet the blogs are proliferating. I, for one, get more useful information and ideas from the blogs than I do from the MSM. I think that the blogs do a very good job of providing information and the error rate is fairly low.
- I know that when someone writes a piece that is not grounded in fact the comments will explode. There are too many smart people in the audience. When writers like me make a mistake it is rare that readers do not bringcall them on the carpet. That was the case with my blog post. The following are some comments from readers:
Just to be clear - of the $284B of investments, only $152B are MBS and of that $152B, only about $47B are private-label MBS. The problem is much smaller but still a potential problem for some banks, such as Seattle, Pittsburgh, Chicago, and Boston.
The comment above is right, you have the numbers completely wrong: the private label mbs are less than 50bln, AND those securities have for the most part already been impaired and written down substantially.
Clearly these folks had a better handle on the facts then I did. I tried to get these individuals to contact me so that I could get the numbers straight. It was my intention to do a clarification but the commentators (no doubt it was someone from the “inside”) did not come forward. If they had I would have done the rewrite.
- I am concerned that this will be used by either the MSM or the government agencies to discredit the blog’s contributions to the public debate on the critical issues we face. It shouldn’t. The proper takeaway should be that bloggers do make mistakes and that readers should always look at the comments to see if there has been a challenge on the facts as presented. There is often more value in the comments than the blog.
- I am well down on the list of bloggers. Less than 15,000 people read the piece in question. But it got the attention of the FHFA and they basically insisted on a correction. My positive spin on this is that the “deciders” are reading the blogs. Possibly we should re-double our efforts.
- The sentence “Fannie and Freddie will cost us at least $400 billion”
was not objected to by the FHFA. I specifically asked them, “Could you confirm that the only objectionable/incorrect aspect of this piece was the statement on the magnitude of loss for FHLB?”
Their argument was with the FHLB loss, not the Fannie and Freddie numbers I used. I am left wondering about that.
- I apologize to the FHFA/FHLBs and the readers for the error. Thanks to the FHFA for bringing it to my attention.