One of the stocks in a list was listed twice - once as "Nippon T&T" and once as "Nippon TT". I have no idea how the author managed to drag that out to so many words, but I know it's a skill I need to learn to make my presentations at work last longer.
I think the author still has a point. Nobody from this "1st class newspaper" has noticed the mistake in 18 months. Their quality control does not seem to be as good as their reputation might suggest.
Even if it’s not used seriously it doesn’t mean it doesn’t have value. Even if it’s nostalgic or an old guy who checked the paper for 50yrs and today still likes checking on a couple of his pet stocks once in while on paper.
This was the thesis of the Emotional Design book (the follow up to famous tea kettle design book) that engineers and amateur designers often over focus purely on function and ignore emotion - they overvalue the most rational approach. For ex: programmers always want the least amount of clicks to do x, when an extra step with an info box and a button might help ease worries and anxieties and smooth the UX.
Often there is value in the pure visceral experience, sometimes emotion, or nostalgia, or pure beauty of the design. This is often what creates a great product that people become religious about. So such things can’t be dismissed out of hand simply because the value cant be reduced to going from point A to B faster/efficiently/cheaper.
Some variant of logical positivism is typically drilled into engineers pretty hard. If it can't be quantified, it's not meaningful. In fairness, this is largely true within the context of engineering-related fields, but generalizes poorly outside of that area. But same way a fish is the last to discover water, it's very hard to discover these assumptions while inundated in them. That "not meaningful" slips by very easily, because nobody stops to ask meaningful in what sense. It already sounds so abstract it's easy to think the line of thought is complete.
Did you not read the article? Most of the post was about how the data being wrong wasn’t the big deal it’s that they shouldnt have the numbers section at all since it means no one reads it.
The fact FT messed up by automating it and not doing basic QA is a signal but it’s not the whole story. A few extra pages for nostalgia reasons, style, and a few hardcore fans is hardly a big burden on a product. But who knows. Maybe it is useless.
Shortly before gitlab was going public I saw a related article where the author called the company github instead. Often I see comments correct the article’s figures.
I still get the WSJ printed paper delivered. Why? No distracting video ads, popovers, cookie warnings, coupon offers, 'subscribe now' pop ups, paywalls, and Taboola clickbait bullshit masquerading as real headlines.
AND I can choose what to read and what to skip, which can’t be done with streaming video. There’s no BREAKING NEWS headlines. There’s no distracting blonde anchor. And if the news is 15 hours old, I don’t care.
There’s nothing in the news that needs my attention within 15 hours after it breaks, except a nuclear missile heading my way.
If I need the news right away, right now, then I read it online.
Do you look up your stock prices in the WSJ tables? Mako's point isn't that print is obsolete, it's that printed daily snapshots of constantly-updated data are.
(Which I 99% agree with, there are some rare cases when you want old snapshots that haven't been subsequently revised, but does that mean we need to keep printing tables in every copy of every day's paper?)
I suspect the average investor would do better focusing on week old stock data that the knew wasn’t being updated than trying to keep up with the market using sub par quotes.
Assuming the goal is to buy and hold for 10+ years, recent changes should be irrelevant.
That's me. I log into my brokerage every few weeks to have a quick peak and make any necessary tweaks. And then I log out and don't think of it. I don't keep an app on my phone because I know it would distract me.
That’s the usual reply I get. No, Adblock doesn’t block everything in my list. It does block many of them, but not all. I’m not asking you to change. If you’re happy the way you consume news, great!
Japanese stocks actually trade with numeric tickers. 9432 is the ticker for NTT. Perhaps there is a bug in the lookup from the name in Japanese characters to the number giving two distinct rows based on whether the name is "Nippon Telephone & Telegraph" or "Nippon Telephone and Telegraph" once translated and then they're using the numeric ticker to look up the underlying data twice.
Not sure it's worth writing an entire article about - bugs in financial data are extremely common. When I worked in equities whenever tehre was a corporate action (split, reverse split, ticker change, merger, dividend etc) in some stock in which we had a position it was always of concern whether the market data providers would make a mistake somehow. Since we had a lot of positions you would tend to get at least one of these errors every few days. Typical mistake would be like, stocks in London typically are actually in GBp (pence) in the marketdata feeds but would often need to be translated to or from GBP (pounds) for some systems. You might see a split or reverse split and the marketdata provider briefly publishing the price off by a factor of 100 for this reason. Now imagine you hold some kind of option or autocallable derivative or whatever. According to the marketdata this should now exercise. However everyone knows this happens and takes care of it. Actually in some contracts we would occasionally see the terms state something like "The price will be the return value of EQ VOD LN HP<GO> on bloomberg at X time" meaning that if bloomberg publishes a mistake you have to run with it.
The amount of effort put into getting reference data correct at big banks is huge, but the volume of data is insane and often there are no good places to verify it.
I had similar experiences when supporting the ref data platform.
On one system holding ETF and Index data we had several thousand products each holding between 10 and a thousand components (eg DJI through Russel 3K). On each product and each component of it there are multiple major fields that need to be correct (price, quantity, multipliers, etc). Any mistake throws off the risk systems and trading in those products (eg basket trading an index, delta hedging, etc)
Millions of data points, and single data errors caused big trading errors.
The data itself could be wrong from the actual providers too, taking debugging into a whole extra level of complexity.
> In the past, the magazine also ran an occasional hidden contest, in which part of the challenge was to find the concealed puzzle with instructions on how to enter (e.g. "You have found the hidden contest. To enter, send us a chain of paper clips."). Until November 2002, readers were also challenged to find the fake advertisement among the legitimate ones (the last one was for the Red Card, a credit card used to pay off credit cards).
At Devoxx, one of the oracle people hid 3 spelling errors in the first slide, the oracle legal disclaimer. Nobody before paid attention to the legal fluff, until that day, when a whole room full of devs scanned every letter of every word.
I wouldn't call this an error. An error would be a wrong number, which is much more insidious. If I actually saw this I'd just look over it and get on with my day.
But I do agree that page is only worth it for the nostalgia.
It's interesting that the author caught this but the premise is completely wrong.
I regular buy the Sunday times in the UK to see what they list the best 5 year mortgage rate as. To the author this looks like a page of numbers that no one reads but that one I care about enough to spend £4.50 on a Sunday times whenever I'm in the UK.
It's often true that there's "a better way to read this on your iPad" but like so many things it's often only "better" if you value your time at nothing.
Sure, a lot of people look online in these post paper days and many are algo driven minute by minute day traders ...
However, to answer the question asked in the article; a not insignificant capital mass of share owners and investors are older than 60 and watch stocks and trades at most day to day, more likely weekly or monthly (sans any sudden news re: MyStock!!).. and "of those people that buy a paper copy of the FT" a goodly chunk have habits that include reading through a paper and glancing at their stocks of interest (alongside tracking general financial industry news).
Probably some of these readers noticed the dupe, most would have written it off as of no importance.
Plus, this data is publically available, on digital format. Anyone who wanted a copy of the data has much easier ways to take it other than copying the outdated version in FT's paper by hand.
Fake symbol and numbers are easy to spot, as they will be crossed with other sources. They are also somewhat risky, as posting literal fake news mixed with real is dangerous.
Doing this is a benign way to see if anyone is sourcing data from you.
The author seems to believe that the fact this error was not reported is an indication that no one reads the printed market data page. That conclusion is not at all supported by the evidence.
First of all, I think everyone is willing to acknowledge that no one reads the market data page line by line from top to bottom. People go to this in order to look up the value of particular stocks and commodities. So anyone who went to this page to look up a different stock (or several different stocks) would never even see this error. Yet they would be benefiting from the FT market data page.
Furthermore, public companies are complicated. For instance: there are two separate ticker symbols: GOOG and GOOGL. On any given day they will have market prices that are slightly different but fairly close to each other. Is this because the Financial Times has accidentally listed this ticker symbol twice? No! It is because Google has chosen to list two different classes of stock on the market. I do not know enough to be sure that listing "Nippon T&T" and "Nippon TT" separately is incorrect -- perhaps one is a subsidiary of the other which, for technical reasons, has the same stock value. I am willing to believe if this author knows otherwise and wants to report this error to the Financial Times. But it is certainly not the case that anyone who had seen this discrepancy would necessarily have reported it, so the continuation of the error is not evidence that no one ever reads this page.
My guess is that the data feed used to have two related shares that became one through a split or other corporate action? And just nobody has bothered to apply a manual correction to remove the dupe.
Whatever the reason I suspect anyone reading that page wouldn't care - there's no incorrect data published after all (as the headline implies)
> My guess is that the data feed used to have two related shares that became one through a split or other corporate action? And just nobody has bothered to apply a manual correction to remove the dupe.
The merge case seems unlikely, as no price would be published for the previous symbol. The multiple share class case seems also unlikely, as that is very seldomly used in Japan.
Most likely the FT tried to create an index of the top 500 worldwide market caps, and got a primary listing of NTT on TSE duped with an ADR or dual listing in EU/US.
I suppose it’s the perils of automation, nobody at the newspaper is going to really spend that much time looking at the output of the script that generates the stock listings page. The stuff written by humans is probably far more in need of proof reading. Somewhere in the chain that autogenerates the page there’s probably a manually edited config file listing the stocks to be included on the page, and someone’s managed to put two entries mapping to TYO: 9432 in it.
They also publish stuff like that. It doesn’t seem to be a very well proofed page. And generally seems to confirm that the whole thing is pretty automated so I can see how this happens.
I could imagine that many readers might casually jump to a few stocks of their interest. If someone were interested in NTT, they might consider a double listing as feature rather than a bug. Two identical consecutive lines might be easier to find quickly (for those born with a pattern-matching inclination).
Perhaps it’s a deliberate error, like the fake towns or roads that mapmakers add to find copycats.
The author is correct though: what a waste of space. I’d actually forgotten that closing prices used to be printed in the paper — I wonder when that stopped?
Yes, it is a mistake, but my inner conspiracy nerd was going, "but what if it is a canary", then i have to tell it, no, that is a terrible canary, just shut up sometimes.
> There’s an argument to made that papers like the FT print these these pages not because they are useful but because doing so is a signal of the publications’ identities as serious financial papers.
If I were reading this paper and these pages subconsciously signaled something to me, I doubt it would be “serious financial paper”. More likely it would be “run by clueless dinosaurs”.
