03 December 2012

Basel 3 and gold

Sprott has a good article out on Basel 3 and gold. Two key quotes:

"It is our understanding that gold’s reference as a “zero percent risk-weighted asset” in the FDIC and BIS literature only applies to gold’s “credit risk” - which makes perfect sense given that gold isn’t anyone’s counterparty and cannot default in any way. Gold still has “market-risk” however, which stems from its price fluctuations, and this results in the bank having to set aside capital in order to hold it."

The "market-risk" aspect is what a lot of commentators who have been saying Basel 3 is great for gold have been missing. If you are not a bullion bank with gold assets already on your books, are you really going to choose gold over cash or bonds when gold is volatile and any fall could wipe out your equity (given their leverage).

I do however take Sprott's point that compared to cash and bonds, gold "provide[s] banks with an asset that actually has the chance to appreciate" and so may be able to compete for space on a bank's balance sheet.

"Basel III will also soon require the application of risk-weights to be applied to the LIQUIDITY profile of both the assets and liabilities held by the bank. ... Under the proposed LIQUIDITY component of Basel III, gold is currently labeled with a 50% liquidity “haircut”, which is the same haircut that is applied to equities and bonds. This implicitly assumes that gold cannot be easily converted into cash in a stressed period, which is exactly the opposite of what we observed during the crisis."

World Gold Council has been working on getting this down but at this time the gold market is still small compared to bonds and equities so at best will have potential only to be a marginal part of a bank's assets IMO.

5 comments:


  1. Bron,

    Sprott writes

    gold’s reference as a “zero percent risk-weighted asset” in the FDIC and BIS literature only applies to gold’s “credit risk”

    As far as I understand it, even unallocated gold would be zero percent risk weighted (although it obviously does involve credit risk). Gold would therefore be on the same footing as any foreign currency (you can hold cash as well as balances at other institutions - cash has no credit risk, but claims on third parties do).

    You write

    World Gold Council has been working on getting this down but at this time the gold market is still small compared to bonds and equities

    Well, according to the Loco London Liquidity Survey, daily trading volume of (paper) gold is more than 2700 tonnes or about $150bn. Thta's already quite a bit, no?

    Victor

    ReplyDelete
  2. $150b is a lot, but not enough I think if you have a lot more giants joining the show. A higher price will solve that a bit by increasing the dollar value of the trade (off same ounces).

    ReplyDelete
  3. Interesting but Sprott underestimates the price risk. Look at 2008. Gold was north of $1000 in the spring and then $700 in October. That’s less than hit that equities took (govies actually rose!), but still very serious. Given that equity in the capital structure of even “solid” banks is under 10% (without even getting into the question of whether asset marks are consistent with reality), giving gold a 0% weight would be reckless.

    As for gold "provide[s] banks with an asset that actually has the chance to appreciate" I would point out that it is not for a bank to make directional bets with its equity buffer. Such bets, if small and confined to a prop desk, are one thing, but a wrong way bet could result in major losses for depositors, customers and ultimately the government. It also creates a perverse incentive system whereby management profits if the bets go well and if not, at worst is out of a job.

    I agree with y’all about gold not having liquidity risk, assuming that it’s not unallocated or derivatives (counterparty risk), which would open the whole CVA desk can of worms, but the price risk is far more important.

    Certainly Sprott and others would like to see another class of bidders pushing up prices, but it wouldn’t be prudent. Also, the whole notion that “AAA-government securities” are not quality assets is off the mark. Treasuries, for example, have zero default risk – the government will always return the money. A valid question is what those dollars will be worth, but since banks are working on a spread it’s a different world. Similarly, Sprott’s notion that this “ultimately pushes government bond yields into negative territory” is off because such yields on the base asset would result in much lower yields on the bank’s assets, e.g. mortgages, preserving the spread. (Plus negative interest rates are very rare.)

    Thus, the flow side, the net interest margin, doesn’t depend on the level of rates. That’s not necessarily true of the stock side, namely the value of assets and liabilities in a changing rate environment. Since banks usually have a positive duration gap, rising rates would eat into the equity cushion. Banks could eliminate the duration gap, but this would be bad for profit, unless the yield curve was inverted. Basel doesn’t adequately address this. What happens if rates rise, lengthening duration on mortgages (bank asset) and shortening on deposits (liability), a double-whammy? Gold (infinite duration?) won’t help the bank (although it might help you!)

    ReplyDelete
  4. Sprott links to, but does not discuss the implications of, the discretionary nature of treating gold as a Tier 1 Asset, proposed by the BIS, per http://www.bis.org/publ/bcbs128b.pdf Footnote 32:

    "However, at national discretion, gold bullion held in own vaults or on an allocated basis to the extent backed by bullion liabilities can be treated as cash and therefore risk-weighted at 0%. In addition, cash items in the process of collection can be risk-weighted at 20%."

    An admirably free-market approach to holding gold bullion, no?

    ReplyDelete
  5. Free market would be getting rid of BIS and all these rules telling you how risky your assets are.

    ReplyDelete