My previous post discussed the price that our new National Asset Management Agency (NAMA) could pay for impaired loans from the perspective of how much of a loss relative to book value the banks could take under the assumption that the government didn’t invest more than its €7 billion planned re-capitalisation. The answer was that the discount from book value would have to pretty small relative to the figures being widely quoted for likely losses.
Admittedly, this was a bit of an around-the-houses way of warming up to the NAMA discussion and Patrick was completely correct in his comment that the key sentence in the speech was
If the crystallisation of losses at any institution requires additional capital the State will insist on participation by way of ordinary shares in the relevant institution.
“Crystallisation” of losses is a somewhat awkward phrase but I’m presuming that the crystallisation occurs at the moment of the transfer of assets from the banks to NAMA at below book value. So the crucial issue is the prices NAMA pays that will dictate these losses. On this issue, the speech doesn’t contain much beyond the following
The potential maximum book value of loans that will be transferred to the Agency is estimated to be in the region of €80 to €90 billion, although the amount paid by the Agency will be significantly less than this to reflect the loss in value of the properties.
One might have hoped that the supporting documentation would have provided more guidance as to how the assets will be valued. However, it doesn’t. Page 6 states
To ensure best value for money for the taxpayer loans will be transferred from the banks to the NAMA at an appropriate written down value depending on an assessment of the value of the loans and the risk being transferred to the State in line with European Commission guidance and subject to appropriate terms and conditions as well as EU State aid approval..
As a first stage, assets should be valued on the basis of their current market value, whenever possible … As a second stage, the value attributed to impaired assets in the context of an asset relief program (the ‘transfer value’) will inevitably be above current market prices in order to achieve the relief effect. To ensure consistency in the assessment of the compatibility of aid, the Commission would consider a transfer value reflecting the underlying long-term economic value (the ‘real economic value’) of the assets, on the basis of underlying cash flows and broader time horizons, an acceptable benchmark
In other words, you should pay market value but, then again, maybe not.
The advice from the ECB in response to the Commission’s guidelines was, if anything, even less promising, in being more explicit about the idea that it was ok to overpay:
The “optimal” degree of public-private partnership should strike a fair balance between risk sharing by banks and not placing an excessive burden on their remaining (“good”) operations … For the participating banks, asset transfers off the balance sheet imply that upfront impairment losses must be recorded, as the transfer prices are lower than current bank valuations. Should banks perceive the impairment losses as substantial, this could reduce their incentive to participate in the schemes or report truthful valuations, at least in cases where a significant part of the bank ownership remains with private shareholders.
So, there we are. We are going to pay prices consistent with EU guidelines and these guidelines are highly (deliberately?) vague.
Interestingly, the supporting NAMA documentation itself provides no discussion of the idea of the government taking ordinary shares. Instead, it puts forward the following:
The State will incur a loss only if the assets transferred to the State cannot over the long term repay the investment made by the State in their purchase from the banks. However, if NAMA make a loss over the long term, the Government intends that a levy should be applied to recoup the shortfall.
To my mind, this post-dated levy is a pretty terrible idea.
Firstly, it leaves a shadow hanging over the banks that we are hoping to have cleaned up and moving on. Secondly, one would have to wonder whether it would ever be introduced. Finally, I would wager a large bet that even if introduced, this levy would only turn out to “recoup the shortfall” in the most narrow accounting sense.
Consider, for example, a levy introduced in 10 years time that pays off the nominal value of the losses slowly over 20 years. In present value terms, if we use a discount rate of B, this stream would be worth a fraction 0.05*(B^10 – B^30) / (1 – B) of the original cost of the funds used to finance this loss. Sticking in B=0.95 (5% annual discount rate), this fraction is 0.384.