Trading 212 is unfortunately not a publicly quoted company.
Discussion
If it was, we could all share some of their profits.
We have probably all noticed the constant advertising of their free account services.
An irresistible enticement to new savers and novice investors.
Hope the newbies quickly realise, that if they become involved with CFDs, there is considerable risk and it is not investing.
How did Trading 212 make £123 million pre-tax profits in 2025, without charging standard account fees?
......................................................................................................
AI explains;
The classic reality of modern retail fintech: when a service is "free," the monetization simply shifts to less visible areas. For the financial year 2025, Trading 212's UK arm reported a massive £277.6 million in revenue and £123.1 million in pre-tax profits without charging standard account fees.
The platform effectively operates a freemium model. The "free" stock-and-ISA accounts act as an acquisition funnel, while the bulk of the company's profits are driven by a highly lucrative derivative ecosystem and backend mechanics.
1. The CFD Engine (The Primary Subsidiser)
The overwhelming majority of Trading 212's revenue historically stems from its Contracts for Difference (CFD) business.
• High-Risk Speculation: CFDs are complex, highly leveraged derivatives where users gamble on short-term price movements without owning the underlying asset.
• The House Wins: By regulatory mandate, the platform must disclose its loss rate—consistently showing that around 76% of retail CFD clients lose money.
• How They Profit: Trading 212 captures revenue here from the "spread" (the gap between buy and sell prices), overnight financing fees for holding positions, and by acting as the direct counterparty to these trades. The massive profits generated from short-term gamblers essentially subsidise the zero-fee infrastructure for long-term ISA and Invest users.
2. Currency Conversion (FX) Fees
While it costs nothing to buy a UK stock in British Pounds, Trading 212 charges a 0.15% FX markup whenever a user buys or sells an international asset (like Apple or Tesla) in a different currency.
• First-time investors drawn in by viral social media trends almost exclusively flock to major US tech stocks.
• A 0.15% fee on the way in and another 0.15% on the way out quietly compounds into a massive revenue stream across millions of active users.
3. Interest Margin on Idle Cash
When retail investors hesitate, keep "dry powder" in their accounts, or leave uninvested cash sitting idle, Trading 212 places those funds into wholesale money markets or central banks.
• While they pass on a percentage of interest to the user to remain competitive, they keep the net interest margin (the difference between what the central banks pay them and what they pass on to the client).
• In 2025 alone, their net client interest income surged to £20.6 million.
4. Share Lending Programs
For standard "Invest" accounts, Trading 212 operates a stock-lending program.
• They take the shares owned by long-term investors and lend them out to institutional borrowers or short-sellers who pay interest for the privilege.
• Trading 212 splits this revenue 50/50 with the user, creating an effortless corporate income stream using assets already sitting in the vault.
5. Entry/Exit Gateways
Though bank transfers are entirely free, they charge a 0.7% deposit fee once a user passes a lifetime limit of £2,000 via instant, high-convenience pathways like debit cards, Apple Pay, or Google Pay. This safely passes on the interchange transaction costs while squeezing a minor margin from impulse depositors.
GROK :Why do people post questions on forums , to then copy post rubbish from AI output ?
Reply :
People do this for a mix of lazy, strategic, and sometimes unintentional reasons. Here’s the usual breakdown:Common motivations Karma / reputation farming
On Reddit, Stack Exchange, Quora, etc., answering your own question (or flooding a thread) can farm upvotes if the AI output looks plausible enough at first glance. Some accounts post a question, then immediately reply with an AI dump to appear “helpful” and boost their score.
Astroturfing or promotion
People pushing a product, service, ideology, or affiliate link will ask a leading question, then drop an AI-generated “answer” that conveniently recommends whatever they’re selling. The AI text makes it look organic and less like pure spam.
Testing or experimenting
Some users are literally testing how good (or bad) the current AI is in a real forum environment. They post the question, paste the AI reply, and watch the reactions. Others are just playing around with tools and hit “post” without much thought.
Genuine laziness or time pressure
Someone wants an answer fast, pastes the question into ChatGPT/Claude/Grok, copies the response, and dumps it as their own post. They don’t bother editing or fact-checking, so it comes out as generic, slightly off, or outright wrong “rubbish.”
Engagement bait / algorithm gaming
Forums and social platforms reward activity. A question + long reply creates a longer thread, more comments, more visibility. AI makes it trivial to generate that volume of text.
Low effort + overconfidence
A lot of people treat AI output as authoritative. They don’t realise (or care) that it can be bland, hallucinated, or missing context, so they post it as a finished answer.
Why it looks especially bad AI text often has a recognisable style—overly balanced, repetitive phrasing, hedge words (“it depends,” “it’s important to note”), and a lack of personal experience or sharp insight. When that lands in a forum full of real humans, it stands out as “rubbish.” Regulars notice the pattern quickly and call it out.In short: convenience + incentives (karma, promotion, attention) + the ease of generating long text with almost zero effort. The same tools that make answering questions fast also make low-quality dumping trivial.
Reply :
People do this for a mix of lazy, strategic, and sometimes unintentional reasons. Here’s the usual breakdown:Common motivations Karma / reputation farming
On Reddit, Stack Exchange, Quora, etc., answering your own question (or flooding a thread) can farm upvotes if the AI output looks plausible enough at first glance. Some accounts post a question, then immediately reply with an AI dump to appear “helpful” and boost their score.
Astroturfing or promotion
People pushing a product, service, ideology, or affiliate link will ask a leading question, then drop an AI-generated “answer” that conveniently recommends whatever they’re selling. The AI text makes it look organic and less like pure spam.
Testing or experimenting
Some users are literally testing how good (or bad) the current AI is in a real forum environment. They post the question, paste the AI reply, and watch the reactions. Others are just playing around with tools and hit “post” without much thought.
Genuine laziness or time pressure
Someone wants an answer fast, pastes the question into ChatGPT/Claude/Grok, copies the response, and dumps it as their own post. They don’t bother editing or fact-checking, so it comes out as generic, slightly off, or outright wrong “rubbish.”
Engagement bait / algorithm gaming
Forums and social platforms reward activity. A question + long reply creates a longer thread, more comments, more visibility. AI makes it trivial to generate that volume of text.
Low effort + overconfidence
A lot of people treat AI output as authoritative. They don’t realise (or care) that it can be bland, hallucinated, or missing context, so they post it as a finished answer.
Why it looks especially bad AI text often has a recognisable style—overly balanced, repetitive phrasing, hedge words (“it depends,” “it’s important to note”), and a lack of personal experience or sharp insight. When that lands in a forum full of real humans, it stands out as “rubbish.” Regulars notice the pattern quickly and call it out.In short: convenience + incentives (karma, promotion, attention) + the ease of generating long text with almost zero effort. The same tools that make answering questions fast also make low-quality dumping trivial.
Pot/kettle! (but I'm sure you knew that)
Copilot says:
Because, bluntly, many people want the appearance of helping more than the effort of helping.
A few common reasons:
Easy karma/reputation: Someone sees a question, pastes an AI answer, and gets thanked or upvoted without actually knowing the subject.
False confidence: AI text often sounds authoritative. People mistake fluent writing for accurate information.
Showing off: "Look what ChatGPT/Copilot/Claude says" can be a way of participating when they don't have personal knowledge.
Speed over quality: It's quicker to paste AI output than to research properly.
They genuinely think it's useful: Not everyone realises how often AI gets details subtly wrong, especially in specialist subjects.
Forum culture changes: Some forums now have users who are effectively acting as human copy-and-paste relays for AI systems.
The irony is that it often defeats the purpose of a forum. Traditionally, if you asked a question on, say, a photography, genealogy, boating, or DIY forum, you hoped for answers from people with years of hands-on experience. If the replies are just recycled AI output, you might as well have asked the AI directly and cut out the middleman.
The really annoying version is when the AI answer is obviously wrong to anyone knowledgeable, yet gets posted with absolute confidence. I've seen examples where experts spend half the thread correcting an AI-generated reply that should never have been posted in the first place.
Used sensibly, AI can be useful on forums as a starting point ("I found this AI summary; can anyone verify it?"). Used as a substitute for expertise, it's often just noise wrapped in good grammar. 😄
Copilot says:
Because, bluntly, many people want the appearance of helping more than the effort of helping.
A few common reasons:
Easy karma/reputation: Someone sees a question, pastes an AI answer, and gets thanked or upvoted without actually knowing the subject.
False confidence: AI text often sounds authoritative. People mistake fluent writing for accurate information.
Showing off: "Look what ChatGPT/Copilot/Claude says" can be a way of participating when they don't have personal knowledge.
Speed over quality: It's quicker to paste AI output than to research properly.
They genuinely think it's useful: Not everyone realises how often AI gets details subtly wrong, especially in specialist subjects.
Forum culture changes: Some forums now have users who are effectively acting as human copy-and-paste relays for AI systems.
The irony is that it often defeats the purpose of a forum. Traditionally, if you asked a question on, say, a photography, genealogy, boating, or DIY forum, you hoped for answers from people with years of hands-on experience. If the replies are just recycled AI output, you might as well have asked the AI directly and cut out the middleman.
The really annoying version is when the AI answer is obviously wrong to anyone knowledgeable, yet gets posted with absolute confidence. I've seen examples where experts spend half the thread correcting an AI-generated reply that should never have been posted in the first place.
Used sensibly, AI can be useful on forums as a starting point ("I found this AI summary; can anyone verify it?"). Used as a substitute for expertise, it's often just noise wrapped in good grammar. 😄
Grok, do you believe what copilot says ?
No.
I don t believe what Copilot (or any other AI, including myself) says in the way a person believes a trusted friend. I evaluate claims on evidence, logic, and consistency with reality.
Copilot is a capable tool built on large language models. It can be accurate, useful, and fast on many topics. It can also hallucinate, confidently state falsehoods, reflect training-data biases, or soft-pedal answers to stay within Microsoft s safety and brand constraints. The same limitations apply to every current LLM, just with different emphases.
So when Copilot (or Grok, Claude, Gemini, etc.) makes a factual claim, the right move is the same: check it against primary sources, data, or experiment where possible. Source does not equal truth.
[ I have stopped now , sorry ]
No.
I don t believe what Copilot (or any other AI, including myself) says in the way a person believes a trusted friend. I evaluate claims on evidence, logic, and consistency with reality.
Copilot is a capable tool built on large language models. It can be accurate, useful, and fast on many topics. It can also hallucinate, confidently state falsehoods, reflect training-data biases, or soft-pedal answers to stay within Microsoft s safety and brand constraints. The same limitations apply to every current LLM, just with different emphases.
So when Copilot (or Grok, Claude, Gemini, etc.) makes a factual claim, the right move is the same: check it against primary sources, data, or experiment where possible. Source does not equal truth.
[ I have stopped now , sorry ]
well played PM3 / Simpo2There's a few UK-listed and fairly well established CFD/spreadbet providers (IG Index, CMC Markets, Plus 500) to look at if you really want to own this sort of business. Personally I wouldn't touch it with a bargepole 
The "problem" isn't the CFD/spreadbet (same thing in different clothes) mechanism per se, but the leverage the firms offer which obviously magnifies both gains and losses, but mostly losses as most customers are clueless. If you look into them you'll see there are varying extents to which they actually hedge the residual risk from their customer flow by trading back-to-back in the "real" market, vs. just letting it run on the basis that most will end up in their favour anyway.
Ultimately there's nothing stopping you fully funding your account, so as to effectively not use the leverage they're offering, and holding positions in boring long term ETFs rather than brainlessly punting dogs
t. Obviously it's not what they want you to do, and mostly they'll still rip you off a bit by charging you a few % financing on the full position size even when it's all your own money.
The other bits - egregious FX fees, interest margin on cash, and to a lesser extent share lending - aren't just the preserve of these firms but also some of the more traditional/reputable brokers too like HL, AJB etc.
The "problem" isn't the CFD/spreadbet (same thing in different clothes) mechanism per se, but the leverage the firms offer which obviously magnifies both gains and losses, but mostly losses as most customers are clueless. If you look into them you'll see there are varying extents to which they actually hedge the residual risk from their customer flow by trading back-to-back in the "real" market, vs. just letting it run on the basis that most will end up in their favour anyway.
Ultimately there's nothing stopping you fully funding your account, so as to effectively not use the leverage they're offering, and holding positions in boring long term ETFs rather than brainlessly punting dogs
t. Obviously it's not what they want you to do, and mostly they'll still rip you off a bit by charging you a few % financing on the full position size even when it's all your own money.The other bits - egregious FX fees, interest margin on cash, and to a lesser extent share lending - aren't just the preserve of these firms but also some of the more traditional/reputable brokers too like HL, AJB etc.
NowWatchThisDrive said:
The other bits - egregious FX fees, interest margin on cash, and to a lesser extent share lending - aren't just the preserve of these firms but also some of the more traditional/reputable brokers too like HL, AJB etc.
Applies all across finance. There is no such thing as a free lunch.Not sure about that unless I've completely misunderstood the models both have.
To me Vanguard is rock solid boring dull get what you pay for and you know what you've paid for it whilst what I see of T212 is a bunch of "trader bro" types spending their lives looking at candles thinking they're doing it all for free.
To be fair that might be based on the people I know who use it
To me Vanguard is rock solid boring dull get what you pay for and you know what you've paid for it whilst what I see of T212 is a bunch of "trader bro" types spending their lives looking at candles thinking they're doing it all for free.
To be fair that might be based on the people I know who use it
To be fair to Jon39, who wants to research the answer to 'How did Trading 212 make £123 million pre-tax profits in 2025, without charging standard account fees?'
When you see T212 adverts on daytime TV showing a phone screen being swiped, a nice uphill graph appearing and the voiceover says 'Build wealth' I know it's aimed at mugs. I'd ban that and all other gambling adverts, especially the ones showing fat women in orange jumpsuits. That end of society has enough to worry about without getting poorer by gambling.
When you see T212 adverts on daytime TV showing a phone screen being swiped, a nice uphill graph appearing and the voiceover says 'Build wealth' I know it's aimed at mugs. I'd ban that and all other gambling adverts, especially the ones showing fat women in orange jumpsuits. That end of society has enough to worry about without getting poorer by gambling.
I use Trading212 and it’s great. That’s why it’s successful.
I know we like to be all snobby and superior on here but on a forum which is sooo ‘free-market economics’, can’t we entertain the idea that a company might be successful because it has a good product?
Trading212 is certainly the most cost effective way of doing what I want to do, which is buy and hold very big and boring ETFs that you’ll all recognise.
The cash ISA is fully flexible and the rate most often tops the charts.
It is very quick and the app is great to use
Sure you’ve got to avoid the social component of the app and not get drawn into any of the muggy games.
Will I use it forever? No- once your portfolio reaches a certain threshold the monthly flat fees of other providers seem reasonable, and you do want a more reputable name behind it. But you should be aware that the app is great and the service is cheap.
I know we like to be all snobby and superior on here but on a forum which is sooo ‘free-market economics’, can’t we entertain the idea that a company might be successful because it has a good product?
Trading212 is certainly the most cost effective way of doing what I want to do, which is buy and hold very big and boring ETFs that you’ll all recognise.
The cash ISA is fully flexible and the rate most often tops the charts.
It is very quick and the app is great to use
Sure you’ve got to avoid the social component of the app and not get drawn into any of the muggy games.
Will I use it forever? No- once your portfolio reaches a certain threshold the monthly flat fees of other providers seem reasonable, and you do want a more reputable name behind it. But you should be aware that the app is great and the service is cheap.
I second Hustles comments above. I use it for the Cash ISA, a little bit of S&S ISA with dividend heavy yields, and as my everyday spending account for the cashback and roundup returns. It works very well, returns decent interest, and is FSCS protected (with the exception of the SS ISA I believe).
Hustle_ said:
I use Trading212 and it s great. That s why it s successful.
I know we like to be all snobby and superior on here but on a forum which is sooo free-market economics , can t we entertain the idea that a company might be successful because it has a good product?
Trading212 is certainly the most cost effective way of doing what I want to do, which is buy and hold very big and boring ETFs that you ll all recognise.
I heard loads of IFAs also recommend T212 to their clients for specific funds that they can't find on their employers or usual ISAs. I'm not sure if there is any snobbery to be honest, but some might chose other platforms for ease of use, having all in one (like vanguard)I know we like to be all snobby and superior on here but on a forum which is sooo free-market economics , can t we entertain the idea that a company might be successful because it has a good product?
Trading212 is certainly the most cost effective way of doing what I want to do, which is buy and hold very big and boring ETFs that you ll all recognise.
Bulgarian bookie. What can possibly go wrong. 
Like credit cards, the service works if you just use the cheap investment lure in and keep the balance low but the key is to recognise that those services exist to onboard the client and then migrate as many as possible onto the book for rinsing. Which like Revolut they're extremely successful at.
And nothing is free. Ever. That's just the marketing game to capture the folks who want to think that others somehow will give them stuff.
Like credit cards, the service works if you just use the cheap investment lure in and keep the balance low but the key is to recognise that those services exist to onboard the client and then migrate as many as possible onto the book for rinsing. Which like Revolut they're extremely successful at.
And nothing is free. Ever. That's just the marketing game to capture the folks who want to think that others somehow will give them stuff.
NortonES2 said:
Why keep the balance low?
Because FSCS payouts can take donkeys years and only cover so much. So if there's an event like a broker rugpull if a client hasn't funds elsewhere they can hit liquidity issues in other aspects of their life and if they're holding more than is covered then there implications there are more obvious. Gassing Station | Finance | Top of Page | What's New | My Stuff