Why Most Stock Tips Lose Money

Every Pakistani investor ends up in a WhatsApp group where someone posts a ticker and a target. Sometimes it works. Over time, for almost everyone, it does not. The reasons are structural, and they are worth understanding — including because some apply to this site.

You are last in the chain

By the time a tip reaches a broadcast group, everyone earlier has already bought. The person who originated it, their friends, the earlier groups — they hold positions. Your buying is what they need to sell into.

This is not always malicious. Information decays as it spreads; enthusiasm is genuine. But the arithmetic is unforgiving: if a tip's value depends on buying before others, and you received it in a broadcast, you did not buy before others.

Thin stocks make it worse

Tips concentrate in small, illiquid companies, because a big liquid name will not move on a rumour.

That is precisely what makes them dangerous. A stock trading a few thousand shares a day can be moved by a modest amount of coordinated buying. The price rises, which looks like confirmation, which brings more buyers. Then the early holders sell, and there is no one on the other side. The stock falls on almost no volume — because you cannot exit thin stocks in a hurry, at any price you like. Before acting on any tip, check the name's average daily turnover and free float — of all the numbers on a stock page, those two decide whether you can get out.

This is the anatomy of a pump. It does not require anyone to be a criminal mastermind. It only requires a thin order book and a group chat.

The shapes this takes in Pakistan

The mechanics above are universal. The packaging is local, and it comes in four recognisable forms.

"Research" from someone who earns on your turnover. Much of the daily commentary on PSX comes from people adjacent to a brokerage. A note saying "accumulate" and a note saying "book profit" both generate a trade, and brokerage revenue is a slice of turnover, not of your gains. Some of it is careful work — but the number of recommendations tracks a commercial need for activity, not how many good ideas exist this week.

"Sure-shot intraday" channels. Telegram and YouTube are full of channels promising same-day calls with entry, stop and target. Intraday is the hardest way to make money in any market, and you also pay commission, spread and whatever tax treatment applies to short-term gains — as of mid-2026, check those with your broker and the FBR. Selling certainty in the hardest arena is not an edge; it is the feeling of one.

The paid signal group. This one is worth understanding properly. The seller's revenue is subscription fees — not a share of your profits, not returns on their own capital. Every incentive follows: fresh calls constantly, so the feed looks alive, and new members arriving faster than disappointed ones leave. A group that told you in January to buy three good businesses and hold them four years would have nothing to post by February. Frequency is the product, churn is the business, and your P&L appears in neither.

The slow walk-up in a small cap. Where the free float is small, a stock can be walked up over weeks on unremarkable volume — a little buying most days, no session dramatic enough to draw attention. By the time the chart reaches you it looks like quiet accumulation, which it is, just not by you. Then the story arrives: an expansion, a contract, a "sector re-rating". The tell is that the move came before the reason. An explanation produced for a rise that already happened is the distribution phase.

Survivorship bias makes tipsters look good

A tipster posts ten calls. Three work well. They screenshot those three.

You never see the seven. Not necessarily deleted — just not mentioned. And because losses are quiet and wins are loud, your impression of that person's record is built entirely from a filtered sample.

Test: ask for every call in date order, timestamped in advance, with entry and exit. Almost nobody can produce it. That inability is the answer.

A target price with no reasoning is a number

"BUY XYZ, target 250" contains no information. Why 250? Based on what earnings, what multiple, what timeframe? Without those, it is a number with a confident tone.

Real analysis shows its work: this company earns this much, comparable companies trade at this multiple, therefore this valuation, and here is what would make me wrong. That last part — what would make me wrong — is the tell. Anyone who cannot state their falsification condition has not analysed anything.

Why the tip feels credible anyway

Knowing all of this does not stop a tip feeling right in the moment. Three things do that work.

It arrives with social proof. Others are already replying, someone posts their fill, someone else is adding more. You are not weighing a claim; you are deciding whether to disagree with a room that has apparently decided.

It is specific. A number resembles knowledge. Vagueness sounds like ignorance, so "target 250" sounds like work has been done — but precision is free. Anyone can type a figure, and the confident type round ones.

Confirmation arrives fast. In a thin stock, the buying the tip produces is itself enough to lift the price. You buy, it ticks up, and the market appears to agree with you inside an hour. What you are watching is your own order, plus everyone else who read the same message, reflected back as evidence. The feedback is real. It is not information.

Nobody is tracking your outcome

The tipster bears no cost when you lose. No fee refunded, no reputation lost — you will not post about it either, because losses are embarrassing. So the feedback loop that would discipline bad advice never closes.

Compare a licensed adviser with regulatory obligations and a paper trail. There is a reason that framework exists, and a reason free tips sit outside it.

How to test a tipster in five minutes

You do not need to investigate anyone. Four questions do most of the work.

1. "Can I see every call you have made, in date order, with timestamps?" Not the wins — everything, including the calls that quietly went nowhere. A record kept in advance is trivial to produce if it exists. The evasion, the change of subject, the offer of more screenshots: that is your answer.
2. "What would make this call wrong?" A real answer names a price, a date or an event — "below 180 the thesis is dead", "if gross margin slips again next quarter I am out". "It will go up unless the market falls" cannot be wrong, which means it was never a forecast.
3. Do the capacity arithmetic. Set the stock's average daily volume against the size of the audience being told to buy. If a name trades 20,000 shares on a normal day and the message went to 5,000 people, most of them cannot get filled near the suggested entry and out near the suggested target — the advice is arithmetically unavailable to the audience receiving it. That disqualifies most broadcast tips in thin stocks.
4. Check whether they are registered to advise. Investment advice is regulated in Pakistan and the SECP maintains registers of licensed entities. As of mid-2026, verify status with the SECP directly rather than trusting a profile bio. Registration is no guarantee of skill. It is a guarantee that someone has an address, an obligation and something to lose.

This applies to us

We publish AI signals. Intellectual honesty requires holding our own output to the standard above.

So: our model is a statistical pattern-matcher over price history, indicators and fundamentals. It does not know news. It has no view on management. It was trained on the past and assumes the future rhymes. We publish out-of-sample accuracy on the model performance page — every prediction, including bad periods, not a highlight reel. For the machinery rather than the marketing, see how these models forecast stock trends.

That number is not spectacular, because honest numbers on hard problems are not spectacular. Anyone claiming 90% accuracy on stock direction is measuring on training data or is not telling the truth.

Our signal is one input. If you take it as an instruction, you have recreated the WhatsApp group with better graphic design.

What works instead

• Understand the business. Slower, less exciting, and the only thing that compounds. Start with the accounts — what P/E, EPS and ROE actually tell you is a weekend's reading that outlasts every group chat.
• Write down your reasoning before buying, including what would prove you wrong. Then check back. Your own record is the only one you can trust.
• Diversify, because you will be wrong sometimes and want to survive it.
• Trade rarely. Costs and taxes compound against you.
• Distrust urgency. "Buy now before it runs" exists to stop you thinking. Good investments survive a night's sleep.

The tip belongs to a family of mistakes retail investors repeat on the PSX — chasing a move, borrowing to chase it faster, averaging down when it turns. The shared root is acting quicker than you can think.

None of this is thrilling. That is roughly the point — the thrill is what is being sold to you, and it is usually the most expensive part.