Boyd Erman of the Globe comments on the HFT study that I reviewed yesterday:
The Bank of Canada has released a research paper on high-frequency trading that, unfortunately for those looking for a silver bullet that finally answers whether HFT is good or bad, provides ammunition for both sides.
…
In other words, it gets harder for those who had been in the market to read the new market, and it’s tougher to trade. That’s a tick in the column of the HFT opponents, who say HFT’s constant use of orders and cancellations to try to figure out the market’s direction creates noise that makes it difficult for other investors.
With respect to the “constant use of orders” point … that’s not what the study says. The study says:
Third, following entry by Aggressive HFT firms, those that mainly take liquidity, incumbents experience a loss in their ability to trade in the direction of future price movements.
…
In contrast, Aggressive HFT strategies are associated with informed trading, since they trade in the direction of future price movements (Biais, Foucault and Moinas 2013; Foucault, Hombert and Rosu 2012; Martinez and Rosu 2013). With more market participants monitoring for arbitrage opportunities, price predictability should decline. We study whether incumbent HFTs are less able to trade in the direction of future price changes.
…
Aggressive HFT entry decreases the price impact of the most informed incumbents’ trades. The average incumbent price impact decreases on average by 0.3 cents for a trade of $10,000. The first Aggressive entries decrease the incumbents’ price impact by a maximum of 1.3 cents.Successive HFT entrants have a diminishing effect on incumbents’ informedness. In fact, there is no statistically significant impact after the second event. This suggests that HFTs are less able to predict future prices and that HFT trades are more reflective of short-term information.
The study does not address in any way the order fill-to-cancel ratio.
With respect to the claim that the actual point is “a tick in the column of the HFT opponents” … I don’t understand it. To the extent that directionality on a five-second time scale exists, what does it matter who gets the money? Is this some kind of argument that it’s better for ‘real money’ to make the profit, as opposed to HFT money? The paper itself refers only to incumbent HFT, not to who was exploiting the information beforehand.
I will make exactly the opposite point, that eliminating directionality on a five-second time horizon is actually a point in favour of HFT. We want markets to be efficient – HFT exploits, and eventually eliminates, the inefficiency on a five-second time-scale. Isn’t that good?
This also reduces the penalty for being poorly informed, which I understand is also a hobby-horse of the anti-HFT mob. In the extreme case, we have a single stock on the exchange, bid at say, 25.00-10, and a single ETF (which holds only the single stock). Ignoring frictional costs, we can then say that ideally the ETF will also be quoted at 25.00-10. But then somebody lifts the offer on the stock with a great big bid and the quote on the stock changes to 25.10-20. Then Granny Oakum puts in her market order to sell the ETF … she only gets 25.00 for it. Isn’t it an objective of market designers to assist Granny to get 25.10? Wouldn’t it be a good thing if somebody (an HFT, for instance) whipped an order to buy the ETF in between the big stock trade and Granny’s order, so that Granny makes an extra dime? Naturally, there will always be the chance that Granny’s order follows so swiftly behind the big stock trade that she doesn’t make that dime anyway. But reducing the time for which this obvious discrepancy is effective is a Good Thing for Granny.
The real problem that the anti-HFT crowd has is that the free dime used to be picked up by a big bank trader with a good pedigree and an expensive private school education. Now it’s going to some bum who nobody’s ever heard of, one of the geeks who was always screwing around with computers instead of attending the polo matches like a normal person. BooHooHoo.
My other complaint about today’s offerings from the Globe is Rob Carrick’s column GICs beat laddered bond ETFs, hands down:
A ladder of guaranteed investment certificates is better, as long as you don’t foresee the need to sell your holdings before maturity. Laddered bond ETFs can be bought and sold any time during the trading days, so they win decisively over GICs on liquidity.
…
If you invest equal amounts in these GICs, thereby creating a five-year ladder, your average yield would be 2.12 per cent.
…
The iShares 1-5 Year Laddered Government Bond Index Fund (CLF) … net yield of 1.22 per cent.
…
The iShares 1-5 Year Laddered Corporate Bond Index (CBO) … after-fee-yield of 1.6 per cent
…
The laddered bond ETF … also offers the possibility of capital gains if interest rates fall, or capital losses if rates rise. Given the flat to rising outlook for rates, losses seem more likely than gains.
So my first objection to this story is in the phrase “you don’t foresee the need to sell your holdings before maturity”. If you don’t foresee that, then why are you holding short-term instruments in the first place? That’s almost certainly just poor portfolio planning. It is possible that you hold the short-term instruments to counter-balance longer-term fixed income (such as the preferred shares that this blog is ostensibly about) – but then you’re frozen into your long-term position as well.
The other objection is “Given the flat to rising outlook for rates, losses seem more likely than gains”, a very common misconception. As I made clear in my article Bond ETFs demystified, there’s no real difference between the two vehicles; the only apparent difference is that an ETF makes visible the opportunity costs of a rise in interest rates that GIC holders like to ignore. Additionally, the objectionable phrase depends upon market timing for its validity and can be ignored solely on those grounds.
The Bank of Canada has released the Bank of Canada Review Spring 2014, with the following articles:
- The Canadian Dollar as a Reserve Currency
- Understanding Platform-Based Digital Currencies
- The Art and Science of Forecasting the Real Price of Oil
- Measuring Uncertainty in Monetary Policy Using Realized and Implied Volatility
- Beyond the Unemployment Rate: Assessing Canadian and U.S. Labour Markets Since the Great Recession
And at the Fed, they’re doing some work on the term premium.
Kevin Carmichael in the Globe comments on Tim Geithner’s book tour:
The European Union now is doing something similar [to the US public stress tests]. We can only wonder what the global economy would be like today if the Europeans had followed Mr. Geithner’s model sooner.
…
The Office of the Superintendent of Financial Institutions does one macro stress test a year and ad hoc tests on specific issues as necessary. But the results remain private, shared only with the Bank of Canada to help it with its twice-a-year assessments of the financial system.Mr. Geithner would disapprove of the secrecy. I didn’t ask him to comment specifically on Canada, but I did ask how important it was that market participants be allowed to see the stress test results. “It’s central,” he said. “You need to let private investors, shareholders and creditors of banks, have enough information that they can better discriminate across institutions. You need to make the loss estimates transparent, you need to make the impact to individual markets transparent, if you are going to allow the markets to provide that form of triage.”
Well, the Europeans eventually did do stress tests, but they were fake, relaxed stress tests, as discussed on July 23, 2010. You only publicize stress tests if you know that the answer is going to be good.
It was a poor day for the Canadian preferred share market, with PerpetualDiscounts off 4bp, FixedResets down 13bp and DeemedRetractibles losing 27bp. The Performance Highlights table is lengthy by recent standards, but balanced with no clear pattern. Volume was above average, with GWO-group issues notable on the Volume Highlights table, presumably due to the GWO new issue.
| HIMIPref™ Preferred Indices These values reflect the December 2008 revision of the HIMIPref™ Indices Values are provisional and are finalized monthly |
|||||||
| Index | Mean Current Yield (at bid) |
Median YTW |
Median Average Trading Value |
Median Mod Dur (YTW) |
Issues | Day’s Perf. | Index Value |
| Ratchet | 0.00 % | 0.00 % | 0 | 0.00 | 0 | 0.4260 % | 2,456.0 |
| FixedFloater | 4.56 % | 3.79 % | 30,644 | 17.85 | 1 | 0.2404 % | 3,768.4 |
| Floater | 2.97 % | 3.12 % | 51,940 | 19.40 | 4 | 0.4260 % | 2,651.8 |
| OpRet | 4.35 % | -6.06 % | 33,696 | 0.14 | 2 | -0.0580 % | 2,710.7 |
| SplitShare | 4.78 % | 4.36 % | 65,696 | 4.17 | 5 | -0.1973 % | 3,102.3 |
| Interest-Bearing | 0.00 % | 0.00 % | 0 | 0.00 | 0 | -0.0580 % | 2,478.7 |
| Perpetual-Premium | 5.50 % | -11.51 % | 97,091 | 0.09 | 15 | 0.0609 % | 2,405.9 |
| Perpetual-Discount | 5.28 % | 5.34 % | 115,448 | 14.92 | 21 | -0.0383 % | 2,550.6 |
| FixedReset | 4.52 % | 3.45 % | 206,050 | 4.27 | 75 | -0.1278 % | 2,562.9 |
| Deemed-Retractible | 4.98 % | -2.42 % | 143,738 | 0.12 | 42 | -0.2724 % | 2,524.5 |
| FloatingReset | 2.65 % | 2.33 % | 143,946 | 4.19 | 6 | -0.0330 % | 2,494.2 |
| Performance Highlights | |||
| Issue | Index | Change | Notes |
| IAG.PR.A | Deemed-Retractible | -1.98 % | YTW SCENARIO Maturity Type : Hard Maturity Maturity Date : 2025-01-31 Maturity Price : 25.00 Evaluated at bid price : 22.74 Bid-YTW : 5.85 % |
| BNA.PR.E | SplitShare | -1.12 % | YTW SCENARIO Maturity Type : Hard Maturity Maturity Date : 2017-12-10 Maturity Price : 25.00 Evaluated at bid price : 25.61 Bid-YTW : 4.40 % |
| IFC.PR.C | FixedReset | -1.10 % | YTW SCENARIO Maturity Type : Call Maturity Date : 2016-09-30 Maturity Price : 25.00 Evaluated at bid price : 26.10 Bid-YTW : 2.51 % |
| PWF.PR.P | FixedReset | -1.06 % | YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 23.98 Evaluated at bid price : 24.31 Bid-YTW : 3.43 % |
| HSE.PR.A | FixedReset | -1.02 % | YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 22.90 Evaluated at bid price : 23.26 Bid-YTW : 3.74 % |
| CIU.PR.C | FixedReset | 1.18 % | YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 21.25 Evaluated at bid price : 21.52 Bid-YTW : 3.54 % |
| BAM.PR.X | FixedReset | 1.46 % | YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 22.41 Evaluated at bid price : 22.98 Bid-YTW : 3.94 % |
| PWF.PR.A | Floater | 2.31 % | YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 19.95 Evaluated at bid price : 19.95 Bid-YTW : 2.63 % |
| Volume Highlights | |||
| Issue | Index | Shares Traded |
Notes |
| TRP.PR.B | FixedReset | 190,737 | Nesbitt crossed 175,000 at 21.00. YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 20.86 Evaluated at bid price : 20.86 Bid-YTW : 3.60 % |
| PWF.PR.P | FixedReset | 127,558 | RBC crossed two blocks of 60,000 each, both at 24.50. YTW SCENARIO Maturity Type : Limit Maturity Maturity Date : 2044-05-13 Maturity Price : 23.98 Evaluated at bid price : 24.31 Bid-YTW : 3.43 % |
| MFC.PR.C | Deemed-Retractible | 109,065 | TD crossed 94,100 at 22.63. YTW SCENARIO Maturity Type : Hard Maturity Maturity Date : 2025-01-31 Maturity Price : 25.00 Evaluated at bid price : 22.41 Bid-YTW : 5.78 % |
| GWO.PR.H | Deemed-Retractible | 107,674 | Nesbitt may have crossed 100,000 at 23.85. The report isn’t clear.
YTW SCENARIO |
| GWO.PR.R | Deemed-Retractible | 95,131 | Nesbitt crossed 80,000 at 23.85. YTW SCENARIO Maturity Type : Hard Maturity Maturity Date : 2025-01-31 Maturity Price : 25.00 Evaluated at bid price : 23.82 Bid-YTW : 5.49 % |
| GWO.PR.P | Deemed-Retractible | 91,052 | Nesbitt crossed 87,000 at 25.85. YTW SCENARIO Maturity Type : Call Maturity Date : 2020-03-31 Maturity Price : 25.25 Evaluated at bid price : 25.69 Bid-YTW : 5.15 % |
| There were 42 other index-included issues trading in excess of 10,000 shares. | |||
| Wide Spread Highlights | ||
| Issue | Index | Quote Data and Yield Notes |
| PWF.PR.P | FixedReset | Quote: 24.31 – 24.63 Spot Rate : 0.3200 Average : 0.2220 YTW SCENARIO |
| CU.PR.E | Perpetual-Discount | Quote: 24.16 – 24.53 Spot Rate : 0.3700 Average : 0.2759 YTW SCENARIO |
| MFC.PR.H | FixedReset | Quote: 26.17 – 26.49 Spot Rate : 0.3200 Average : 0.2268 YTW SCENARIO |
| ENB.PR.T | FixedReset | Quote: 24.36 – 24.65 Spot Rate : 0.2900 Average : 0.1977 YTW SCENARIO |
| CU.PR.F | Perpetual-Discount | Quote: 22.21 – 22.65 Spot Rate : 0.4400 Average : 0.3556 YTW SCENARIO |
| IGM.PR.B | Perpetual-Premium | Quote: 26.00 – 26.24 Spot Rate : 0.2400 Average : 0.1596 YTW SCENARIO |
Another thing Rob Carrick ALWAYS misses is taxes. Bond ETFs are stuffed with premium bonds which have adverse tax consequences in taxable accounts. Basically, the premium disappears upon maturity, creating a capital loss, which only gets half the tax back that was paid on the premium coupon (assuming tax losses are useable — offset against gains somehow). As a result, the effective tax rate on a Bond ETF these days is between 50 and 75% — and that is after MER! Some GOC short term bonds actually have negative after-tax yields (i.e. effective tax rates >100%). The ETF doesn’t care if those bonds are in the index — it has to buy them.
So, take XSB the I-shares short term bond ETF as of May 14, 2014: NAV $28.70, Average Term 2.83 years; Coupon 3.01%, Yield to Maturity 1.55% and Duration 2.68. A simple bond price calculator in Excel yields an average bond price of $104.02. After 28 bp of MER, the pre-tax YTM is 1.24%. After tax, assuming 46% marginal tax rate AND the ability to use the $4.02 capital loss in 2.83 years, the after-tax return is 0.35% (it is 0.24% if the investor can’t use the capital loss, and sadly, many can’t). 0.35% after tax on an investment yielding 1.24% gross is a 71% effective tax rate. Better to keep the money in a high interest short term bank account! a 3 Year GIC at 2% yields 1.08% after tax — nearly 3 times as much.
Every time I bring this to Rob’s attention he pooh-poohs the problem with a statement something like “most people invest in Bond ETFs in their RRSP”. That way he can forget about taxes again.
One good thing about GICs is that they are par bonds, so the effective tax rate is your marginal tax rate. Strips have the same advantage.
Taxable investors are probably better off purchasing bonds individually and searching out bonds with prices near or below par.
Quite right. It’s a major problem in the current environment. I have several consultation clients who have to jump through hoops to get a suitable fixed income portfolio in their taxable accounts.