Another AI ‘discussion’ — more-or-less slop-free, as far as I can tell.
It prompted some additional human intelligence thoughts. Obviously, the $535m Solyndra loss was due to credit default, whereas the $2.1b loss at EPA WIFIA was a funding loss. But there are some significant similarities:
- Solyndra was a loan guarantee — that is, a federal commitment to fund an obligation under specific circumstances regardless of losses. Likewise, WIFIA made $25b of loan of fixed-rate loan commitments that could be optionally drawn at the borrower’s request for many years after the commitment date. Of course, WIFIA’s highly rated borrowers requested $10b in disbursements when circumstance were most favorable to them — and loss-making for taxpayers.
- In both cases, the expected cost to taxpayers at commitment was a fraction of the total loan volume involved. In Solyndra, the $535m guarantee was apportioned $78m in credit subsidy, or 14.5% of the amount, for expected credit losses. For the WIFIA portfolio, the credit subsidy estimates for the entire $25b in loan commitments was a trivial $185m (0.75%), reflecting the very high credit quality of the loans.
- At least Solyndra’s apportionment was focused on the right risk — credit default — even though the amount was wholly inadequate. In WIFIA’s case, OMB missed the funding risk entirely and apportioned nothing for predictable funding losses, despite the requirements of their own credit policy Circulars. Maybe Solyndra’s true credit risk was predictable — WIFIA’s funding losses certainly were. Both were failures of due diligence — arguably, WIFIA’s was worse.
- In both cases, what was pitched as enabling a real-world policy outcome turned out to be a pointless transfer payment. Solyndra’s entire business model failed — there was no ‘policy return’ for taxpayers at all, even though Solyndra employees, contractors, suppliers, etc. did in effect receive federal money. For WIFIA, to the extent the loan commitment actually accelerated infrastructure investment (doubtful), disbursement at a loss to taxpayers added nothing to that — by definition, construction was already underway. The ‘windfall’ coming from cheap WIFIA financing was at that point simply a transfer payment from federal taxpayers to select water ratepayers, regardless of need. In other words, even if there was a ‘policy return’ for taxpayers on the $185m of discretionary funding apportioned to the WIFIA portfolio for expected credit losses, there was zero return on the $2.1b of funding losses.
The real question in this comparison, however, is not about such substantive matters, but rather about the role of ‘narrative’. For Solyndra, narrative drove both the commitment and the intensity of the subsequent scandal. For WIFIA, the narrative has always camouflaged what was actually happening at the Program (i.e., the ‘critical investment’ story instead of the ‘free interest rate option’ reality).
But what will the narrative be for the funding loss consequences? Silence? Spin about ‘meaningless accounting abstractions’ or something like that? What are the political dynamics? In the context of upcoming reauthorization, it’ll be interesting to see.