Member Info for taufour

"Still a chunky seller."

Diagramatically ... see luku.org/xi/BMN/84jrjj9b/ ... that flurry of four chunky (150k shr) sells between 12:30 and 1:15 is a definite feature,

but they are decisively overshadowed during the rest of the afternoon by 5 trades in the 200k shr range which look like 4 buys and 1 sell

Those articles are so good that you might want to correct even the tiniest slips ...

where it says "Monday, 19th Mar 2019" ( at https://www.thebushveldperspective.com/blog/public-trade-analysis-6/post/trading-analysis-19th-march-2019-417 ) note that "monday" should be "tuesday"

"... the signficant number of small trades are thought to originate from the freetrade platform ... there is nothing frictionless or high-frequency trading about it"

Interesting ... and maybe testable!

If the very small trades are done by institution-style frictionless algorithms, one would expect them to be closer to the midpoint price than most other trades.

If they are done through consumer-style apps, one would expect them to be further from the midpoint price than most other trades.

Does that make sense?

Thanks. Maybe I haven't had experience enough of the dark side of the market to really appreciate what you guys are saying. I do wonder at the large number of very small trades but have so far been assuming that there is an innocent explanation for them.

I've been imagining that many or most of them are done by practitioners of frictionless trading ... where the cost of a trade is just the electricity running their computers ... and they have algorithms constantly twitching things around in microscopic steps driven by a vast number of exotic rules

like for example, if the cost of rubber in Malaysia goes up by a hundredth of a percent then sell a hundred dollars worth of vanadium shares

But that as I say is just imagination, and I suppose not very plausible.

But what other explanations are possible? I am mulling over your (Jimbo) suggestion about "the large number of tiny trades that currently get used to artificially widen the spread and skew the data." Could you explain how that would work exactly? What is the point of artificially widening the spread?

Aaah ... it's the weekend again, and I have an idea.

The spread plotter at luku.org/tau got stuck last November because the data-feed which it was using for getting MM quotes ceased operation. I tried looking for another publicly accessible free machine-readable data-feed for MM quotes and didn't find any.

But does it really need MM quotes?

In ideal market conditions, the ask and bid lines form a gently undulating envelope, within which trades snake along in two bands, an upper band and a lower band.

So how about, instead of using MM quotes to draw ask and bid lines, simply trying to discern the two bands directly, and drawing lines along the top of the upper band and the bottom of the lower band. Better not call them 'bid' and 'ask' because those terms refer to MM quotes, which we do not have. Instead we could simply say that the upper and lower lines indicate entry and exit zones.

Someone please tell me if this is lunacy. I'm thinking of giving it a try.

(Nov 26th) luku.org/xday/ "... still has the flaw that it can double-report trades ... I will try to fix this next weekend and let the forum know if I do"Fixed, I hope. Also, it is lining up columns better.:-)

Hi, just need to slip this post in before tomorrow's no-posting day.

Jimbo, "unfortunately it doesn't retrofit trades reported at a later date (like today's big ones) back to the appropriate date..."

Thank you for letting me know. I've improved it a bit so that if you run it without filling in a date, it fills in today's date and runs the cron job there and then, thus responding with the latest available data, which should include yesterday's after-hours trades.

I think that eventually all late-reported trades get fed in to the historical data but I should check.

It still has the flaw that it can double-report trades (because the stock exchange has a propensity for amending its listings through the day to improve accuracy - for example it can fill milliseconds into the time-stamp field, or amend a price of "12.34" to "12.3395")

I will try to fix this next weekend and let the forum know if I do.

Hi Jimbo. The cron job should pick up any trades reported up to the time it runs, which is late in the evenings.

For previous days' data, if you don't mind typing in the date backwards (yyyy-mm-dd), there is http://luku.org/xday/

Hi faramog: "on Sharesmagazine, only the first transaction ... is listed ... I downloaded the current record an they were not on it .. neither were a couple of others on your list"

As a rough estimate from daily logs, I think that on average only about 90% of the BMN trades published by London Stock Exchange make it on to the list published by Shares Magazine.

Feeding the data into a visualization tool can reveal interesting patters. BMN today so far, http://luku.org/snap/BMN/dtzawrc5/

"I dont think this is just transfer of stock as the sells wouldnt have been broken up into smaller chunks as we see here."

Just guessing here, but I think it still could be a transfer of stock, bearing in mind that there could be some reason related to their taxation regime making it necessary for them to do the transfer on the open market rather than off-market, and therefore making them break it into smallish chunks so as to help keep the market stable. But as I say, this is only guessing ... opinions from accountants would be welcome.

Yes. Did exactly that ... but only with a very small part of my holding. Just to stave off boredom.

Hello folks. I'm curious about whether that "email" from crux in the first post of this thread could possibly be genuine.

I have no reason to doubt the bona fides of the poster (Paddyboy13). I'm dubious about the substance of the post.

I would discount the spelling mistakes etc ... evidently the poster did not copy-and-paste but carelessly transcribed.

But what seems odd is this: "...they will be reaching out with new date,potentially around some fresh news..."

Even though companies obviously do come out with fresh news periodically, I would have thought an outfit like crux would not say what they are quoted as saying, even if it's true.

Shares magazine article about "Mining winners and losers from Tesla's Battery Day"

Ramifications for: Glencore (heavily into both cobalt, down, and nickel, up)

Biggest winner: Vale

Possible upticks for: BHP, Anglo American

Might benefit in time: Horizonte

No other companies mentioned (apart from Tesla of course)

How come I was able to buy on the bid, smack in the middle of a long apparently calm trading period when the quoted spread has not changed a jot?

Bid/ask has been steady at 12.50/12.75 all afternoon.

There's been a fair bunch of trades, mostly at 12.50, with a sprinkling of obvious buys above 12.67 (and with just 5 trades in an unclear middle range around 12.58)

The only outlier was the 400k whopper at 12.4063 shortly after lunch, flagged PRIC+LRGS

Well the odd thing is that at least one of the trades at 12.50 was a buy (mine, 32k shrs timed at 16:03).

Alfacomp: "I'd be interested to know how your Cron job gets its information - there's a lot of statistical analysis that could be done on the back of that."

Hi, I've sent email.

Meanwhile, for general interest, here is a web page that does the same thing as the cronjob:

luku.org/xlon/

Hope it is useful and/or fun.

Actually it doesn't seem to like these long posts full of numbers, that was less than 50 lines.

It's quite easy to do this, if anyone is interested I'll be happy to post code (but where? Not sure if this forum would put up with it)

51 trades of 500k:

2020-06-25 11:07:19.213 2020-06-25 12:53:38 GBX 14.45 500000 72250.00 Off-Book AIMX LRGS
2020-07-15 13:37:54.767 ---------- --:--:-- GBX 13 500000 65000 Off-Book AIMX
2020-07-15 13:47:39.673 2020-07-15 14:40:49 GBX 12.8 500000 64000.0 Off-Book AIMX LRGS
2020-07-16 08:15:47.663 2020-07-16 09:14:18 GBX 12.1 500000 60500.0 Off-Book AIMX LRGS
2020-07-16 10:14:20.370 2020-07-16 11:13:53 GBX 12.275 500000 61375.000 Off-Book AIMX LRGS
2020-07-20 09:57:19.147 2020-07-20 10:54:42 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-07-20 11:11:08.090 2020-07-20 12:10:20 GBX 11.5 500000 57500.0 Off-Book AIMX LRGS
2020-07-21 10:15:32.287 2020-07-21 11:14:47 GBX 11 500000 55000 Off-Book AIMX LRGS
2020-07-24 08:11:51.373 2020-07-24 09:08:45 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-07-27 07:44:59.427 2020-07-27 08:44:16 GBX 12 500000 60000 Off-Book AIMX LRGS
2020-07-28 07:46:48.143 2020-07-28 08:43:33 GBX 11.575 500000 57875.000 Off-Book AIMX LRGS
2020-07-28 10:39:07.327 2020-07-28 11:34:18 GBX 11.575 500000 57875.000 Off-Book AIMX LRGS
2020-07-28 15:23:58.933 2020-07-28 15:35:13 GBX 11.8 500000 59000.0 Off-Book AIMX LRGS
2020-07-29 09:33:07.097 2020-07-29 10:30:20 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-07-29 13:52:21.977 2020-07-29 14:49:55 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-07-31 07:38:07.117 2020-07-31 08:37:56 GBX 11.3 500000 56500.0 Off-Book AIMX LRGS
2020-08-03 09:14:13.490 2020-08-03 10:13:31 GBX 11.35 500000 56750.00 Off-Book AIMX LRGS
2020-08-03 11:33:39.127 2020-08-03 12:25:59 GBX 11.35 500000 56750.00 Off-Book AIMX LRGS
2020-08-03 12:46:51.943 2020-08-03 13:44:59 GBX 11.4 500000 57000.0 Off-Book AIMX LRGS
2020-08-04 07:06:57.453 2020-08-04 08:04:35 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-08-04 07:54:02.180 2020-08-04 08:53:31 GBX 11.6 500000 58000.0 Off-Book AIMX LRGS
2020-08-04 09:21:17.627 2020-08-04 10:19:55 GBX 12 500000 60000 Off-Book AIMX LRGS
2020-08-06 07:45:40.427 2020-08-06 08:45:34 GBX 13.25 500000 66250.00 Off-Book AIMX LRGS
2020-08-06 12:20:00.000 2020-08-06 13:10:04 GBX 12.765 500000 63825.000 Off-Book AIMX LRGS
2020-08-11 07:32:22.353 2020-08-11 09:11:04 GBX 12.625 500000 63125.000 Off-Book AIMX AMND+LRGS
2020-08-12 08:33:54.377 2020-08-12 09:28:20 GBX 13.1 500000 65500.0 Off-Book AIMX LRGS
2020-08-14 09:52:30.000 2020-08-14 10:44:35 GBX 12.6 500000 63000.0 Off-Book AIMX LRGS
2020-08-20 10:42:26.000 2020-08-20 12:38:15 GBX 14.15 500000 70750.00 Off-Book AIMX LRGS
2020-08-26 11:15:25.067 2020-08-26 12:13:38 GBX 13.3 500000 66500.0 Off-Book AIMX LRGS
2020-08-28 09:12:28.370 2020-08-28 10:11:38 GBX 13 500000 65000 Off-Book AIMX LRGS
2020-09-01 12:14:42.770 2020-09-01 13:14:00 GBX 12.5 500000 62500.0 Off-Book AIMX LRGS
2020-09-01 12:23:02.680 2020-09-01 13:22:06 GBX 12.6 500000 63000.0 Off-Book AIMX LRGS
2020-09-01 14:55:26.160 2020-09-01 15:35:57 GBX 12 500000 60000 Off-Book AIMX LRGS
2020-09-01 15:11:57.860 2020-09-01 15:35:57 GBX 12.5 500000 62500.0

Bummerd: "Does anybody count the 500k drops through time?"
and Daisan: "There are also 1m and 250k sells that I would include."

The following may be of interest, from a cronjob that updates daily from XLON. Since June 4th it logged 3 trades of 1m shares, 51 trades of 500k shares and 50 trades of 250k shares (one of which is a UT). Times are in GMT, the second time if given indicates delayed publication time, and buy/sell indication is not included.

3 trades of 1m:

2020-09-08 13:32:38.037 2020-09-08 16:06:59 GBX 12.5 1000000 125000.0 Off-Book AIMX LRGS
2020-09-09 08:25:42.000 2020-09-09 15:26:33 GBX 12.75 1000000 127500.00 Off-Book SINT LRGS
2020-09-11 08:54:01.000 2020-09-11 16:08:17 GBX 13.625 1000000 136250.000 Off-Book SINT LRGS

I'll put the 500k trades and 250k trades in separate posts because there seems to be a line limit.

xerx: "14k sold 41k bought yet sp eases down"

I think MMs have found that the best way to prod PIs into selling is by lowering the price.