14 Responses to “May 31, 2022”

1. cbh says:

   June 1, 2022 at 10:13 am

   Thank you all for the great commentary we have seen here recently. Today, I would like to introduce a new subject that others may wish to comment on.

   For the last few months I have seen a noticeable increase in programmed trading in Canadian preferred shares. One very clear indication of this is the instantaneous reaction to any new bid or ask that is submitted. The programmed trading is unpleasantly aggressive. For example, if I put a very modest bid in for say 200 shares, instantly I am outbid by one penny for 2500 shares. Another case is when you see an existing bid of 100 shares, and you put a bid at the same price. In this case, the 100 share bid moves up one penny, because it is actually an iceberg order and they will not tolerate an equal bid.

   Having been a programmer in a previous life (now retired), I was instantly interested in how smart these programmed trades were. They were astonishingly stupid as of a month or two ago. If I was continually outbid, I could slowly raise a 100 share bid and push their programmed bid up to high prices. I could then cancel my bid, and the programmed bid would stay at a foolishly high level. Within an hour or so the programmed bid would get filled, and sometimes the problem would go away (or, sometimes we would start the same process over again).

   Within the last month or so, the programmed trading is unfortunately getting a bit smarter. You can still push the bid up so that the programmed trading gets the worst price of the day, but you generally must leave your own 100 share bid in there to make that happen.

   Perhaps I should not be so vindictive here. However, professional interest makes it enjoyable for me to explore the evolution of their trading algorithms. The people behind the programmed trading also deserve some punishment, because they are not willing to play nice. Do they have to insist on outbidding your 100 share bid, when they are trying to get 5000 or 10000 shares?

   My take: when you run into this overly aggressive, but naive programmed trading, do your best to make it costly for them!

2. skeptical says:

   June 1, 2022 at 10:24 am

   cbh:
   Thanks for this lovely post. Those who actively trade the market are familiar with this programmatic trading and it’s almost certain that any bid you have for any issue is likely to get a penny increment.
   How to beat them?
   Wait for the down markets. Wait for the volatile markets.
   Last month when the preferreds were in the depths of despair, such bids had disappeared from most issues. But the aggression is back in play now.

3. skeptical says:

   June 1, 2022 at 10:47 am

   And on a different note, it seems bonds and stocks are once again moving in the same direction.

4. skeptical says:

   June 1, 2022 at 11:10 am

   And the flat pancake is here.
   5 year GoC=2.88
   30year GoC =2.88

   another couple of hikes will almost do the trick.
   The US is fairing a whole lot better with still a mildly positive slope between 5 and 30.

5. baffled says:

   June 1, 2022 at 11:41 am

   cbh , thanks for that . i sell lots of naked puts and i notice the same thing with my option orders .

6. paradon says:

   June 1, 2022 at 11:59 am

   Programmed trading as you described has been around for a few years now. I find it is mostly with a handful of issues that trade frequently and are investment grade. Still lots of opportunity with outliers though!

7. ratchetrick says:

   June 1, 2022 at 12:45 pm

   I find exploring the algorithms a good way to get a better price if you’re actually looking to sell. In this case, I generally start with a buy order a penny above the bid. If it’s immediately jumped, I start moving up my bid . . . and I continue to move it up until the algorithm hits its outer high bid, at which time it usually retreats to it’s original bid price. If his high bid is acceptable, a quick move back up to that point (less a penny), brings the high bid back . . . I then hit it with my sell order, kill my strategic buy order . . . and move on.

   I think the guys who set up these program buy/sells are probably the same guys that spend $40 on a Big Mac meal to be delivered to their door lol! Like John Belushi quoted in Animal House, “fat and stupid is no way to go through life!” . . . but I’m glad they’re out there rotflmao

8. paradon says:

   June 1, 2022 at 1:12 pm

   Well since I am one of those guys I don’t agree with your characterization!

   Basically an application calculates FMV for a particular security at a particular time. Different people do it to varying degrees of complexity. In my case if the bid drops a delta below FMV I place an order at bid plus $.01. If you walk me up a few cents I don’t consider that a loss. I still purchased the security a significant discount to what I felt it was worth.

9. ratchetrick says:

   June 1, 2022 at 1:28 pm

   I guess the point would be that FMV on most pref issues, being highly illiquid, and being quoted with spreads that are often too wide to be taken seriously, is not really possible to establish reliably. Considering as well, that moves in the bond market, central bank banter, etc. can have an immediate impact on James’ next day selection of rainbow vs bomb blast pictorials, it would appear that the last idea that would make sense would be to try to design an algorithm to confirm FMV on a pref.

   paradon, cudos to your technical originality, and desire to participate in the market in this fashion . . . and perhaps it’s worked out for you. That’s wonderful . . . I assure you though, investors that are willing to participate “in person”, as opposed to virtually, will have far more success reacting to the real factors that influence pricing in this category. Just a few thoughts!

10. stusclues says:

    June 1, 2022 at 4:15 pm

    “I guess the point would be that FMV on most pref issues, being highly illiquid, and being quoted with spreads that are often too wide to be taken seriously, is not really possible to establish reliably.”

    Any buyer can calculate his/her/their own FMV for any issue (if by that what is meant is a price that provides the buyer with a sufficient/fair future return based on his/her/their own assumptions). Bid/ask spreads are irrelevant. They only matter to execution.

11. paradon says:

    June 1, 2022 at 5:45 pm

    Our goals are different ratchetrick. I use programmed trading to find mispriced issues and buy/sell them. I am not buying to hold in general.

12. skeptical says:

    June 1, 2022 at 5:53 pm

    I use programmed trading to find mispriced issues and buy/sell them.
    For novices like me, what platform do you use for market data APIs?
    What about programming? Which language do you write your programs in and which trading platform lets you execute these trades?

    Thanks

13. paradon says:

    June 2, 2022 at 10:25 am

    I use Questrade’s API. It is older, but there are code samples on GIT to get you going. I program in C#.

    Lots of things I don’t like about their API but their data coverage is good and they are generous in terms of how often they allow you to poll their API even with the free tiers.

14. skeptical says:

    June 2, 2022 at 10:43 am

    Thanks paradon. Appreciate the insight.