Another bout of massive stock market volatility. Another day of frustrated private investors glaring at their frozen screens like horny teenagers trying to download a low-res porn MPEG on a dial-up modem in 1994.
When markets get super congested like they did on Monday, retail platforms fall over. It doesn’t matter whether we’re in the midst of a crash like we saw back in March or if shares are going gangbusters as with this week’s vaccine rally. If you’re a private investor trying to buy or sell shares, you’ll be lucky if you can log into your broker, let alone trade.
You’d hope headlines like these would focus minds at the platforms:
- Investors rage as market surge crashes trading platforms – CityWire
- Hargreaves Lansdown suffers system outage amid record trading volumes – FT
- Retail trading hits snags as vaccine news sparks stock scramble – Reuters
But this problem is hardly new, so maybe their strategy is just to grin and bear it…
…until the next day of pandemonium rolls around.
Musical shares
If you’re a dedicated passive investor – good for you – then you may say “so what?” to this kerfuffle.
Passive investors don’t trade like hyperactive card sharps. Passive players buy, sell, and rebalance their holdings according to their long-term plan. Ideally they automate the whole process. They would then be oblivious to the disruption their active brethren endured earlier this week.
At Monevator, we certainly believe most such investors who use broad index funds will do better than those who try to beat the market.
They’ll also sleep better at night!
I tried to tell a friend about Monday’s market mania. My friend has learned his passive investing habits on this very website, from my passively pure co-blogger. My friend was bemused, because his portfolio just appeared to have gently risen a couple of percent since the weekend. And he’d only looked at it because I asked him to.
To prove I wasn’t an overly sensitive soul, I sent him some commentary on the market rally, such as this snippet quoted by Bloomberg:
Based on historical data, the book-to-market [factor] enjoyed a 12x standard deviation rally, while price momentum and short-term growth factors suffered from 20x and 25x sigma sell-offs, respectively, on November 9.
That’s a lot of sigmas.
Most of the indices just marched higher. But I own technology shares that fell 20-25% over Monday and Tuesday, as well as value stocks that gained that much and more.
This churn is what market wonks (guilty as charged) call ‘internal rotation’.
Partly it’s reflective of a change in sentiment among investors about the earnings outlook for different sectors.
This time around we saw ‘stay at home’ tech stocks made less appealing by the positive vaccine news, and concurrently more appetite for burned-out ‘physical economy’ firms that need boots on the ground to make money.
Yields on safe government bonds ticked higher, too. If that continues it would be bad for high-multiple shares priced on their long-term earning potential, as I explained a few years ago. At the same time certain beaten-up value shares such as banks could profit from higher rate expectations.
Yet another driver of internal rotation is when traders sell one sector as a source of funds to buy shares in another.
Even if an investor has spare cash, using it to buy the suddenly more appealing airlines, hotels, and cinema chains would mean increasing overall equity risk. Whereas selling other shares at the same time tamps down your overall exposure.
Well, it does if you’re actually able to access the market via your platform.
Going for broke
I use multiple brokers and they nearly all gave me trouble on Monday.
For a while I couldn’t even log into one. Others would let me in, but then they wouldn’t let me trade.
For example, I got into Freetrade instantly with my thumbprint and it showed me my holdings without breaking a sweat. I was all set to sing the praises of its shiny modern tech stack – until I tried to actually buy some shares. Multiple attempts left me waiting for a buy confirmation that never came – the trades were never executed.
Other brokers appeared to execute my live trades but then, after a long timeout, they admitted that really they couldn’t even get a quote from the market.
Far worse, there are reports of investors on some platforms buying more shares with phantom cash that was never deducted from their balances after earlier trades executed, and even of big negative cash balances once the dust had settled.
The platforms should have been able to uncross all this – but not without infuriating customers, and possibly dinging their potential profits.
Interactive Investor appears to have held up best among the major platforms, judging from social media. That’s interesting given it has not always received the most sparkling marks for customer service. Perhaps it’s been investing in technical capacity instead of phone lines?
Price sensitive punters
Should we care that so many platforms fell over when the market went crazy? Should we be angry customers?
I think you can be miffed about it while still acknowledging the platforms have a difficult job.
People always say that Internet-enabled businesses should be more ready for any sudden surge in demand – online grocers, video aggregators, and share platforms alike.
And of course they mostly are prepared. But what level of extraordinary extra demand is it reasonable to cater for?
You can be ready for a market that’s 10-times busier than average. But then it will be the 11-times busier market that will get you every time.
I do think these platforms are different from other sites that fail with demand surges, though. It doesn’t really matter if you order your granola from an online supermarket with a 15-minute delay. In contrast share prices change constantly, and access to those prices is exactly what you’re paying for.
You could also argue a very active stock market is clearly one that many investors want to be, by definition, given all the activity. So if a platform fails to enable you to get involved, is it even fit for purpose?
To reiterate, at Monevator we think most people should be passive investors. They should turn off their PCs and smartphones on a day like Monday and go for a walk. Try to pick short-term winners and losers during such a feeding frenzy and you’re liable to lose a limb. Or at least a few quid.
Even so, it’s not very credible to argue that a platform failing on a very busy day is protecting small investors from themselves.
You could equally well say a wine producer watering down its alcohol or a cigarette maker stuffing its fags with parsley is doing consumers a favour.
Good luck getting that past Trading Standards!
Play to play
Set against all this moaning, owning a DIY portfolio has never been cheaper or easier. Low-cost platforms are a huge reason why.
We might clamour for a quant-fund’s fat pipe plugged into a market that’s gurning like a clubber in 1980s Ibiza, but would we pay for it?
The very biggest platforms are making decent profits, so you might argue they can afford to upgrade their infrastructure.
But it’s also true that the last time the regulator looked deeply into this sector a few years ago, the rest of the platforms were making diddly-squat.
There’s been consolidation since then, so the situation may have improved. But it would be counter to investors’ interests if pressure for bombproof platforms – perhaps even from the authorities – led to more mergers, less competition, and with that higher prices. Be careful what you wish for!
Did you try to buy or sell shares earlier this week? How did you get on?
Note: We both get a free share if you sign-up via my link to Freetrade. The Interactive Investor link is an affiliate link, too. The author owns shares in Hargreaves Lansdown.
Comments on this entry are closed.
What I worry about is that platform servers will somehow forget how many units I own – do they keep multiple backups in different places save from solar flares and EMPs and power surges and computer viruses and whatnot? Would they detect a data corruption and restore the good copy or would they fail to notice and end up backing up the corrupted data overwriting the good copy?
I imagine the system is too big to ever be allowed to fail that way, or there’d be hell to pay, but I wouldn’t put it past them when the likes of TSB melted down and many banks use legacy computers. I worry every time there’s “maintenance work” or computer system upgrades
Mentioned this on another forum, I had a look at the tech behind HL
A curious mix of new and old with the indication that they might be using their own server farms.
As such they won’t have the ability to autoscale to more compute in case of demand spikes.
I could write a IT strategy for them but they might flinch at my 0.45% fee, but I do have really good customer service.
I work for a mobile phone co.
Demand peaks are a real problem, even when they are known about in advance. If you put the infrastructure in place to deal with massive surges happening maybe once a year, you end up with a massively under-utilised (on average) network with associated costs -which would be passed on to the customer.
Investors that use “cheap as chips” investment platforms and then wonder how it is that they cannot trade in real time in a stormy stockmarket need to get real
Your asset allocation should be set to ride these market ups and downs
Most market movements up and down -but of course generally up -occur over a very few days
Being out of the market trying to trade with one of these platforms will lead to severe losses
If you want to day trade get a stockbroker and pay the high fees
Otherwise trade as little as possible -if at all-and then only in the “quiet” times .
xxd09
@BerkshirePat
Auto-scaling has been a thing in the cloud for a decade. You don’t have to provision infrastructure up front for demand spikes, only to make the systems do that when needed. If only more retail brokers were using the cloud…
Re: Auto-scaling, I was a bit disappointed with Freetrade in this respect. But on reflection, perhaps its problem (and that of some other brokers?) was with an intermediary further down the line — the market maker providing the quotes or whatnot.
Perhaps I could interview a couple of retail platform providers and see what they say, if people think there are interesting questions to ask?
@Slavo:
AFAICT, the cloud is not infinite. Thus, there will always be constraints, it is just a question of where they lie and what compromises you are prepared to take/make to manage them.
Re savings platform profit. Alliance Trust was held back for years by its trading platform ( since sold to II). Suspect scale is everything noting the way II is hoovering other platforms up, so margins for small players are probably thin.
This article speaks volumes about the lack of NFRs particularly performance-testing as well as the need for Cloud-based auto-scaling (and backup) solutions, that will allow “modern” platforms to attract more customers towards investing.
Equally, I like the Vanguard approach of “No Mobile apps – You do not need to be checking your investments that quickly” 🙂
Wasn’t bothered too much about the issues, but agreed that a certain standard is expected with newer technology companies(or companies purporting the use of newer technologies).
> Are retail share dealing platforms fit for purpose?
Normally I apply the rule that any headline couched as a question is a) not worth reading and b) the answer is no, but since this is Monevator I excepted that rule 😉
Yes, they are perfectly fine. The clue is in ‘retail’. I used to own a retail jigsaw, made out of pressed metal. It couldn’t saw true, and the gearbox stripped after a couple of years.
I looked at the one in a builder’s toolkit. Cast footplate, well beaten up, but presumably still serviceable. I looked up how much it cost, and I wasn’t prepared to pay that much. I am an amateur – he is a professional.
If you’re going to be getting into the markets during times of major volatility, you need professional tools, and presumably you have enough AUM or an on a high enough screw to justify the costs. You aren’t on a scummy nominee account where you have “flay[ed] costs as if they were the tattooed agents of darkness”
You need the right tools for the job. Professional investor – you or your employer pays for the necessary gear. Retail investor – accept your tools will fail you in service under abnormal loads.
It costs real money to have capacity. I too worry about #1’s concern of data integrity, but that is a very different matter from a system being able to handle a massive peak to mean ratio. If you work of Goldman Sachs and your platform goes titsup then you and GS are entitled to hop up and down spitting bricks. Retail investors, not so much. We’re not prepared to pay for a big capacity crest factor, so guess what? We don’t get it!
In the days before the internet, how would a day like Monday have played out?
Jammed phone lines to brokers to ‘sell’ or ‘buy’? Who would investors have raged at as they listened to that infuriating engaged bleeping tone?
But maybe the phone lines wouldn’t have been so jammed since not many people had a hotline to their own brokers back then?
The internet has made investing easily accessible to many folk but it looks like the technology might still be playing catch up to cope with the growth, and as mentioned, needs to adapt, eg move more onto the Cloud to cope with these surges when they happen.
As somebody who works in and with cloud suppliers, as Al Cam pointed out, they are finite. A large London DC belonging to Azure stopped all new devices being spun up at the height of the lock down due to resourcing issues. Another London DC experienced a UPS failure and was out for “days”. So they are fallible.
I think the best advice is from xxd09 : “Otherwise trade as little as possible -if at all-and then only in the “quiet” times “.
As a realist and technologist, I think Vanguards view on mobile access is refreshing, “no” there isnt as app for that, still chuckling at that….
Auto-scaling in the cloud is the key but bear in mind that’s no silver bullet either.
a) It takes some time before new instances come up so that you can handle the load
b) Newer platforms like FreeTrade can’t do everything themselves. They depend on 3rd parties – can’t remember the US company allowing them to trade there. The bottleneck was probably just moved further down the line…
c) Sometimes even cloud platforms fail. See FreeTrade’s outage on 6th of Oct because of a DNS issue in Google cloud platform (although being multi-region would definitely help!)
All things considered, I would trust a newer platform over an older one for sure. Migrating tech is hard which is why legacy platforms have such a high cost of maintenance and roll out features a lot slower.
Is it more likely that these brokers are just ignoring ‘the little people’ and putting all their resources into the big boys at times like these?
When you read the autobiography of the owner of Renaissance Technologies-a certain Mr Clements-you realise what the amateur investor is up against
This is a very successful hedge fund-churns out millions of pounds of profits every year
Been doing it for many years
He employs only the top Maths PhDs ,the most up to date technology-pays a fortune for it and his computers trade in real time-ie millisecond trades etc etc
Constant trading is a Losers Game for the average investor
xxd09
Well, I did nothing as it never occurred to me that there would be something to do. Or I was busy. Or something.
@Ermine – I agree. I pay as little as possible to meet my requirements. The ability to trade a broad range of mainstream UK funds and etfs is about it. My budget is more Timex wind-up than Rolex Daytona. And ii meets that splendidly and even includes a trade every month in the price!
iWeb offers a price-limited trading service called Tradeplan which I had set to sell an iShares ETF when it hit a new high and the screen showed this trade, set weeks earlier, as ‘Executing’ for more than half an hour compared to the usual few seconds. The site was very slow for the rest of the day. But it did do what was required, so I cannot complain.
12spacebadger99, Al Cam, Foxy
Of course nothing is infinite. Scaling comes at the cost of maintainability and complexity. But Cloud at least gives you the choice to make that tradeoff yourself. I’m also a technologist and have the fair share of war stories with cloud.
I was just providing a counter-point to @BerkshirePat – you don’t need pre-provisioned infrastructure to handle demand. It’s a choice these days, not a must-have.
Regarding outages – it’s easy to blame “cloud” for them, but not all are because of cloud no matter how many brokers use it as an excuse.
@Slavo totally agree!
I was recording my Oct divs on Mon which all went fine in the morning…then when I went back in the afternoon I couldn’t access AJBell or iWeb. Fair enough, the vaccine news which I thought was an inevitible event proved to be market breaking. But going back two weeks I couldn’t even get a quote to buy MRCH (negotiated trade only with iWeb) even when volatility was relatively low. I set up a trade plan only to find that the trade failed to execute despite several trades executing above my trade plan price. So no, some of the retail platforms are not fit for purpose. Would I want to pay more? Probably not. But having some broker diversity definitely helps in times of strife.
@Slavo Point taken re auto-scaling- not my area really, but sadly not possible for mobile networks! Assume there could be other bottlenecks in the broker systems?
Until I read it here I did not even know that had been problems. My attention was elsewhere this week! Had I wanted to trade though I would have found it annoying, but if there is extreme volatility in the market I would usually want to wait until that settled down before I traded.
I doubt if the platforms run on third party cloud computers as that would be difficult from a regulatory point of view. I would be surprised if it made any difference if they did anyway as the bottlenecks were unlikely to have been due to a lack of CPU or memory resources. More likely database IO, lock contention or network bandwith to intermediaries and/or exchanges would be the issue. You cannot solve all IT performance issues by chucking more CPUs at them.
I care far more about competence than I do about odd days when a platform may be a little slow.
I jumped in the market no trouble using int investor. Managed to buy a few more shares for my heavily deflated portfolio . As usual, I’m down again.
If you’re interested in the lengths the professionals will go to to get that extra speed, check out ‘Flash Boys’ by Michael Lewis
I like @ermines analogy. I had a similar experience with drills. I bought what I felt was an expensive one, retail. Was not willing to spend any more. It is OK for some things but try drilling into the outer walls of my house…. Had to borrow a builders drill to get the job done – using the same drill bits of course. But I drill so infrequently can’t justify the expense of anything more powerful.
When you get websites falling over someone always thunders “BUT THE CLOUD” in disbelief. The cloud i.e. such as AWS and Azure services are just giving you additional on-demand server capacity. That’s all they do. However that’s not always what the problem is. There can be all kinds other bottlenecks in systems e.g. if you are relying on external companies to do some part of the trade execution then you are at their mercy, or if some part of your design doesn’t scale particularly easily no amount of servers will magically fix it.
Yeah, I think Ermine’s builder’s tools analogy is spot on. I’m not spending a fortune on either tools or investment platforms as I’m neither a professional investor or builder. If I was, I’d get the expensive ones that hold up under heavy use.
@ermine, richard, matt:
Are any of the retail power tools in question mains?
IMO batteries (or more accurately cells) is another area of so-called progress!
> Are any of the retail power tools in question mains?
The jigsaw was. But it was only £50 ISTR. Buy cheap buy twice. I don’t have a jigsaw any more. Similar quality to a typical retail platform IMO 😉 The builder’s one was well north of £200, and cut straight both in the vertical plane and along a line. Somewhere I read that you should really call the ones with a cast as opposed to a pressed guide a totally different tool, I think that fellow was right.
I refuse to buy any more battery tools because:
* I don’t have off-grid land any more, so I can easily get mains power anywhere I want to use it. It’s more inconvenient to use each time, but that’s not as inconvenient as having to replace power tools every three years because the battery has died
* batteries are evil because they change the connections every two years, just to stiff you, and the pricing is like printer ink – the profit is in the batteries
I have tried taking the batteries apart and repacking them, but you can’t identify quality or not in Chinese replacements, and the cost of locally sourced ones is such that you are better off replacing the entire thing. Having worked as an electronic engineer means I don’t wreck batteries as often as most folk. Any sort of Ni-whatever chemistry, recharge the damn thing as soon as you hear the drill run slow from lack of voltage, because the weakest cell is about to get back-charged at a high rate. Which is exceedingly bad for it. I am still running a Makita cordless drill on the two original batteries from 2013, though they are very tired now. On the farm with general punters using the same model of drill batteries got stuffed in six months.
I’m reasonably happy with my mains power tools. The jigsaw has been only the second mains tool I’ve had that failed in service. I did manage to wreck a Black & Decker mains drill when segments of the commutator started to fall apart. But it had been my Dad’s in the early 1970s and was nearly 50 years old when it hung up its boots, and I’d been using it for 25 years.
Random detour into DIY tech aside, the principle still stands. If you want performance under adverse circumstances, you have to pay for it. That is not commensurate with flaying costs as if they were the tattooed agents of darkness. You don’t get ‘owt for nowt. Much of the thrust of this site is that retail investors shouldn’t be charging in and out of the markets on the back of the news feed. As a retail investor, if you have to do that sort of thing then spread betting can perform better under such conditions because a) it is a model of the markets, and gapping can manage instantaneous changes (ie spreadbetting follows the markets accurately, except under adverse conditions where tracking errors and spreads increase) and b) it is marketed specifically to people who want to trade frenetically.
Earlier this year I did not try to use my retail trading platform (iWeb and Charles Stanley). I shorted some of my own holdings using IG index. It worked out for me, although it did effectively transfer money out of my ISA into the SB account. Obviously the official advice is here be dragons, don’t try this at home. But if you don’t have tens of millions of assets under management, and you are vain enough to think you want to swim in that pool, then a little bit of lateral thinking will get you a service that meets your occasional needs. Just bear in mind that the vast majority of people that use spread-betting platforms lose money hand over fist, and the cost of carry is very high, these are not for any long term positions at all.
Couldn’t log on to Iweb for most of that day. I didn’t want to trade, rather to see how much of my paper losses had temporarily reduced! But as the article says, some unable to trade will have lost out, others would have benefited. Day traders were disadvantaged yes. However, there’s plenty of domestic gains likely to arise in coming weeks, with the Pfizer full results and the Oxford results. If both are as hoped. Reverse it true too. That am, just before the Pfizer results came thru, I bought my first ever global technology tracker on CSD. Hey presto the Pfizer results reversed the momentum away from tech. My golden rule: whatever you do, it’ll often be the wrong call, at least immediately!
I use Barclays Smart Investor, never had any issues – all works well including this week.
Doesn’t appears in Best Buy lists but they have a price improver that reduces the buy / sell spread which I like, think they link into the Barclays Investment Bank infrastructure
I didn’t know about the problems until I read about it in the news because I rarely trade, other than my monthly drip feed after pay day. I update my spreadsheet at the start of the month, which is when I get my latest portfolio figure.
I’d be annoyed if my platform spends a huge amount to improve performance to cope with abnormal days and increases my charges accordingly. Sorry to be selfish, but I don’t want to pay for an “improved” service I don’t need: that’s not “treating customers fairly”.
Youinvest, Hargreaves, etc. are surely mass-market products aiming to keep costs low and as far as I know, don’t claim to provide a high performance trading system with service guarantees (at premium cost).
In addition, I wonder how much the problems were exacerbated by new, inexperienced people “panic” buying i.e. the same people panic selling in March when Covid became serious in the U.K.? There seems to be more first time investors than ever and I suspect they haven’t yet learnt the hard way that passive is better… and on that note, I shall leave.
Keep up the good work Mr Monevator!
Stay safe everyone
FT has a story up with a bit more detail:
https://www.ft.com/content/dd3cf99a-ff71-44d4-831c-0edac79d3104
Re. Stockmarket ups and downs, I think I learned my lesson after the Financial Crisis 2007-2009. I hadn’t discovered self-select ISAs and had a simple Halifax Sharebuilder into which I had a D.D. of about 200 quid going in every month, and used The Motley Fool to help decide on which shares -all in UK – after which I never looked for ages. Working abroad, I rarely watched the news and had a lot of stress at that time with work and family, so the Crisis was already well under way by the time I noticed. So I did nothing. Anyway, I had no time to do anything. A few years later, everything was back to normal and my holdings all zooming upwards. The only changes I made were to read up on Asset Allocation and start investing in ITs and ETFs to access the rest of the world. What had been a 100% UK single companies (mainly high yielding) portfolio slowly morphed into Equities (high yielding AND growth), Fixed income, Commodities, Property and Cash, with the UK accounting for approx 30% of all that – and still going down! I have made a small change since Covid started and that is to channel more cash into fixed term savings, at least until I see the pandemic coming to an end.
Some of the most intelligent posts I’ve seen in responses to a thread. Completely agree with @ermine’s analogy and the biggest take away from this whole article was advice on recharging early to prolong battery life.
@theinvestor – ‘market that’s gurning like a clubber in 1980s Ibiza’ – that’s the quote of the week without a doubt. and one I’m going to be using somewhere at some point.
IT side of things – unless you control EVERYTHING (probably impossible) and have complete redundancy again on everything (unlikely), you are always going to be relying on someone’s else’s capacity/ability to deal with bottlenecks at some point. Which may not be tested to the levels’ you want/expect or suddenly appear out of the blue (Monday). However, the failings of the platforms the day after Brexit is another matter – that should have been considered and planned for accordingly.
> After Monday’s meltdown, the FCA said: “We expect investment platforms to plan for unexpected scenarios . . . and maintain operational resilience.”
OMG FCA get your damn sweaty mitts off our retail platforms! I’ve already taken the shaft from the RDR making unit trusts dearer for me to hold because they stopped the kickbacks, which means I have to pay a cost of carry on regular investing because the kickbacks subsidised my platform fees by parting the unaware from their money. Now they want to jack up the costs of these platforms by demanding they deliver Bloomberg performance in market turmoil. Noooooooo please leave well alone!
Well-informed punters buy a retail platform because for the vast majority of the time they want to flay costs as if they were the tattooed agents of darkness. The reason is that they want the lowest costs of carry. You’re sitting on your backside waiting for the market’s puny 3-4% p.a. return lift the aggregate value most of the time. You trade infrequently, you can tolerate paying as you trade, because most days you are passive, what part of this word do people not get, passive means doing nothing! You don’t want to pay platform fees for doing nothing.
If you want to trade like a mad thing, then pay for performance. Passive investors don’t need performance. They need cheap. If you can’t fiddle with your portfolio while the market is going bananas, that’s not such a bad thing. We are little people guys. We can’t afford to pay for platforms with Goldman Sachs performance, because fees will go up and eat all our returns. It costs money to provide resilience and cope with a high potential demand. It’s perfectly OK for our retail platforms to go titsup when the market is having a hissy fit. At the moment you have the choice – cheap with low fees but you have to ride turmoil out, or high fees and you get performance.
FCA – leave our low-cost rubbish performance platforms alone. We don’t want to pay for performance with a higher cost of carry, because guess what? We’d be using something better than a cheap platform if we did. Users have the choice of better performance right now. Pay for it yourself if you want it, don’t make us cheapskates pay over the odds. By all means make platforms specify we don’t expect our system to handle a sudden 10x increase in load without falling over, but that’s as far as it needs to go.
I’ve not even a passive investor, but this talk gives me the willies, because I am low-churn enough that this would hit my returns. I don’t want to pay for the infrastructure costs of coping with high transaction rates. Gerrroff my cheap and nasty platforms FCA!