« What next for the Bank of Canada? | Main | The perils of pricing publicly-provided products »


Feed You can follow this conversation by subscribing to the comment feed for this post.


Your interpretation of the Royal Bank statement in the ROB is correct I think.

But the problem is not that a minority of banks’ funding costs are tied to the overnight rate. It’s that a majority of their asset revenues are. (Majority and minority used in comparative terms here.) So there is a mismatch between assets and liabilities that are tied to the overnight rate – more assets than liabilities. This is referred to as an “asset sensitive” interest rate mismatch. When rates drop, revenues decline more than costs. Technically, the sensitivity isn’t to the overnight rate per se. It’s to the prime rate and administered rates that tend to move with the prime rate, as well as short term market rates that tend to correlate with the Bank of Canada rate. Contributors to the problem on the asset side include prime rate tied business loans, variable rate mortgages, and large pools of liquid assets that the banks routinely hold. These are all sensitive to a decline in short rates. You can see the banks now fighting the problem on the variable rate mortgage side because they no longer offer negotiated discounts from posted rates – rather premiums. Another challenge is that the Canadian banks have huge fixed rate mortgage portfolios that are exposed to declining rates because of prepayment and refinancing options. Rates come down and people want to get out of higher rate mortgage costs if they can. Fast prepayments on mortgages reduce the duration of the portfolio – known as the “negative convexity” problem for mortgage portfolios.

Also interesting is the contribution of equity capital on the other side of the balance sheet to the whole interest rate risk problem. Royal Bank for example has about $ 30 billion in common share capital and retained earnings. There’s no interest rate payable on this capital – in other words, it’s a fixed zero rate of interest. They have to figure out what sort of interest rate behavior will offset this on the asset side. If it’s variable rate, falling short rates will compress margins on this part of their balance sheet. (This is apart from the fact that they’re expected to pay a dividend on their capital, a dividend that results from margins and profits on their entire balance sheet, not just this section.)

Deflation and falling rates are a potentially big problem for banking system interest margins, one of the unintended consequences of central bank easing in response to credit risk problems.

JKH: I hoped I would see your thoughts on this post!

"It’s to the prime rate and administered rates that tend to move with the prime rate, as well as short term market rates that tend to correlate with the Bank of Canada rate."

But, why does the prime rate tend to move with the Bank of Canada overnight rate? If we think of the prime rate as the price of the banks' output, if banks' costs move less than one-for-one with the overnight rate, why would the prime rate?

I think of the prime rate as being chosen by the banks themselves, as the price of output in a monopolistically competitive industry.


The prime rate is an “administered” rate as opposed to a market rate. The central bank rate is also an administered rate. Bank rate changes both lag and lead associated market rate changes. A bank rate change lags to the extent it was “anticipated” by the market, and it leads to the extent it was not (e.g. market uncertainty over 25 or 50 basis points would roughly split the lag and lead effects if the bank went 50). And of course the bank rate normally leads the prime rate.

Administered rates move in step function whereas market rates move continuously. You’d have to ask a historian as to how the administered rate function of banks came into being. But it makes sense for several reasons I can think of. One is that it’s still a stabilizing force for bank interest margins, compared to an institution that is driven entirely by market rates, such as an investment bank. Two is that the data processing and human capital requirements required to convert an administered rate system to a real time market rate system would be overwhelming and unnecessary from a retail customer perspective. The typical customer doesn’t need live quotes on a prime rate or mortgage rate that is changing on an intra-day basis. Three, the effect of a stable administered rate is to average out spreads between that rate and a corresponding market rate over time (e.g. prime rate versus the short term cost of funds). The net effect is to strip out market volatility from what would have been a long term average anyway, had the administered rate been converted to a real time market rate, incorporating the same credit spread. Administered rates are a way of looking through pure volatility of markets rates (which is unproductive) in setting bank pricing.

The result of all this is that there is an interaction of two sections in a bank balance sheet – the administered rate portion and the market rate portion. The administered rate portion contributes a fairly stable margin whereas the market rate portion has a margin that changes in real time. Banks pay close attention obviously to the combined effect of these two pieces in managing their balance sheet risk. They have a pretty good handle on exactly how a particular prime rate change will affect their margins, which have already been affected by market rate changes leading up to the bank rate change. Approaching the zero bound, with general margin compression at risk, banks obviously have to weigh the competitive and PR considerations of a prime rate change that may actually hurt their margins to a significant degree.

JKH: The way I think of it, the prime rate of (say) TD Bank is administered by TD in the same way that menu prices are administered by a restaurant. The restaurant faces a downward-sloping demand curve for its meals. If it raises the menu price, it loses some customers (to other restaurants, or to eating at home), but not all. Similarly if TD raises its prime rate, it loses some customers (to other banks, or to not getting a loan at all), but not all.

So both the banking industry and the restaurant industry is monopolistically competitive. Their menu prices/prime rates are set by the restaurants/banks, and this is a market process, but it is not a perfectly competitive market, and the prices change in a lumpy rather than continuous fashion.

Raw food is only a small percentage of restaurant costs. If raw food prices came down by x%, I would expect restaurant menu prices to come down, but by less than x%. And I would not expect restaurants to have lower profits as a result of a fall in raw food prices. Same with banks, and a reduction in the overnight rate.

But if restaurant customers insisted that an x% fall in raw food prices translate into an x% fall in menu prices, I can see why restaurants wouldn't be happy if raw food prices fell.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad