Pump and dump and penny stocks: how the scam works

Basics10 min read · updated 2 September 2026 · 11 sources

In a pump and dump, fraudsters buy shares cheaply, drive up the price with promotion and rumours and sell to the buyers they have lured in. Then the price collapses. The preferred targets are penny stocks: small shares with a low price, little trading volume and little information. Here you will learn how the scheme works, who wins and who loses, what the law and research say about it and which warning signs help you recognise it. This article explains a fraud; it is not a how-to guide.

What penny stocks are

In Germany, “penny stock” is not a defined legal term. Colloquially, it means shares with a very low price (often a few cents) and a small market value (penny stock). According to Leuz et al. (2017), pump-and-dump schemes almost always revolve around such shares. Most are traded over the counter or in less regulated segments, such as the open market (Freiverkehr) in Germany, AIM in London or OTC Markets in the US, formerly the Pink Sheets. In the sample of German cases, the median market value of the promoted companies was just over €11 million; they were often shell companies without any significant business.

For fraudsters, such stocks have three advantages: low liquidity (even small purchases move the price), few mandatory disclosures, because fewer rules apply on the open market than on the regulated market, and little scrutiny from analysts and the press. Anyone who spots a mispricing can hardly exploit it for lack of liquidity, for example through short selling.

How the scheme works

1. Building a position: the fraudsters quietly buy large quantities at low prices. Sham trades among themselves (wash trades) can fake rising prices and lively trading in advance.
2. Promotion (pump): stock newsletters, spam emails, phone calls, social media posts and messenger groups promise a multiple of the stake, often with a big story about commodities, patents or takeovers. In the German cases in Leuz et al. (2017), 37% of campaigns used email, 32% newsletters and 31% the telephone.
3. Price rise: new buyers drive the price up; the rising price itself becomes a selling point.
4. Sell-off (dump): while the campaign is still running, the fraudsters sell to the buyers in tranches. Repeated spam on consecutive days keeps demand high and extends the window for selling (Hanke and Hauser 2008).
5. Crash: when the promotion ends, the buyers disappear. The price falls steeply, and low volume makes it hard to get out.

Worked example: who wins, who loses?

A made-up case with round numbers shows how gains and losses are distributed. It is meant as an explanation, not a template: pump and dump is prohibited in the EU and a criminal offence in Germany.

The buyers pay the fraudsters' profit almost one for one. Real cases are usually less extreme, but they go in the same direction: in the sample of Leuz et al. (2017), investors who took part lost almost 30% on average. According to Frieder and Zittrain (2006), those who bought on the day of the heaviest spam promotion and sold two days after it ended lost just under 5.5% on average before fees. A spammer who bought heavily promoted stocks the day before and sold on the day of the heaviest promotion earned 4.29% on average before costs.

Variants: short and distort, crypto pumps, scalping

• Short and distort: the reverse. Fraudsters sell a stock short and then spread false negative reports in order to buy it back more cheaply. BaFin describes such short attacks in its own publications (BaFin 2018).
• Crypto pumps: according to Kamps and Kleinberg (2018), the pattern in cryptocurrencies has spread to shorter time frames and more markets. Li et al. (2025) document numerous pumps with drastic swings in price, volume and volatility and a rapid reversal; the fact that prices already rise before the start points to wealth transfers from outsiders to insiders.
• Scalping by influencers: someone with access to the media buys a stock, then recommends it publicly and sells into the price rise without disclosing the conflict of interest. This is exactly what Article 12(2)(d) of the MAR names as market manipulation.
• Distinction: with a meme stock, widespread enthusiasm in online forums drives the price, not necessarily a fraudster intending to sell. It is risky all the same.

Legal framework: the MAR and the WpHG

An overview, not legal advice. In the EU, the Market Abuse Regulation applies (MAR 2014). Its Article 12 defines what market manipulation is, including: transactions and orders that give false or misleading signals about supply, demand or price, or that secure an artificial price level (paragraph 1(a)); the dissemination of false or misleading information and rumours through the media, including the internet, where the person disseminating it knew or ought to have known that it was false (paragraph 1(c)); and voicing opinions in the media about a financial instrument after having taken a position in it, without disclosing the conflict of interest (paragraph 2(d)). Article 15 prohibits market manipulation and attempted market manipulation.

In Germany, under the Securities Trading Act (WpHG § 119 1998), intentional market manipulation that actually affects the stock exchange or market price is punishable by imprisonment of up to five years or a fine, or by one to ten years if committed on a commercial basis or as a gang; attempts are punishable too. Below that threshold, market manipulation, even when committed recklessly, is an administrative offence punishable by a fine (BaFin 2018). BaFin publicly warns about ongoing campaigns, for example about stock tips in WhatsApp groups. This works: in the data of Leuz et al. (2017), around 25% fewer investors bought the promoted stock after a public warning (21 cases), above all ordinary investors, while very active traders were hardly affected.

What the research shows

• Who manipulates: Aggarwal and Wu (2006) analysed enforcement cases of the US Securities and Exchange Commission (SEC). Manipulators are usually well-informed parties such as insiders or brokers. During the manipulation, prices, liquidity and volatility rise; afterwards, prices fall.
• Spam works: Böhme and Holz (2006) found clear short-term reactions of volume and price to spam campaigns. In Frieder and Zittrain (2006), the probability that a promoted stock was the most traded in the sample rose from 4% on days without promotion to 70% on days with promotion. According to Hanke and Hauser (2008), spam increases excess returns, volume and intraday volatility without a lasting price effect; liquidity is an important factor in its success.
• Social media: according to Renault (2017), unusually large numbers of Twitter messages about small stocks go hand in hand with a strong price rise on the same day and a sharp reversal in the following week; network analyses reveal clusters of suspicious accounts such as promoters, fake profiles and automated posts.
• Who falls for it: Leuz et al. (2017) studied 470 schemes and the brokerage accounts of more than 110,000 clients of a large German bank. Almost 8% of active retail investors took part in at least one scheme and lost almost 30% on average. Not all of them were victims in the narrow sense: some day traders deliberately speculate along. Portfolio composition and past trading behaviour explained participation better than demographic characteristics.
• Crypto: according to Li et al. (2025), pumps reduce the liquidity of cryptocurrencies and depress their prices; a taste for gambling, overconfidence and naive learning from past successes explain participation.

For advanced readers: detecting manipulation in data

Most studies are event studies: around the day of the promotion, they measure the excess return relative to a benchmark and the abnormal volume. A rise before or on the promotion day followed by negative returns is typical. There are pitfalls, however:

• Selection: only campaigns that someone has collected are captured (spam traps, enforcement cases, conspicuous accounts). Undetected manipulation is missing.
• Cause and effect: promotion often follows a price rise that the fraudsters themselves triggered with advance purchases; Frieder and Zittrain (2006) found positive returns already on the days before heavy promotion.
• Theory: according to the model of Aggarwal and Wu (2006), it is precisely investors who seek information about the true value who make manipulation easier: the more of them there are, the greater the competition for the shares and the easier it is for manipulators to trade.
• Detection: Kamps and Kleinberg (2018) use anomaly detection methods to search for unusual phases in price and volume data and found clusters on certain trading venues and coins.

Warning signs

• Unsolicited recommendations by email, phone call, messenger or direct message.
• Promises of a rapid multiplication of the share price, time pressure, “today only”.
• A very small company with a low share price on the open market, without verifiable revenue but with a big story: a commodity discovery, a miracle cure, a breakthrough technology.
• A sudden multiple of the usual trading volume without any announcement from the company.
• Newsletters or influencers who do not disclose whether they hold shares themselves or are being paid.
• The rising price as the only argument.

Limitations and common misconceptions

• Not every price jump is fraud: genuine news, takeovers or widespread enthusiasm also drive small stocks strongly.
• Channels change: many spam studies date from 2006 to 2008; today, campaigns also use social media and messengers. The mechanics remain the same.
• Getting out in time? If that is your plan, you are betting against fraudsters who decide the timing. In the worked example, they are already selling during the rise.
• Only detected cases: case collections show what has come to light; nobody knows exactly how common manipulation really is.

This article is not legal or investment advice. Related topics: speculative bubbles, insider trading, alternative data and stock picking.

How Portfolio Lab calculates it

Portfolio Lab does not analyse penny stocks and does not detect manipulation. The data checks look for faulty price series: for stocks, they flag daily moves of 50% or more, and for ETFs and funds, moves of 25% or more. A single outlier day with a log return of more than 0.35 (about +42%) that reverses to within 0.10 on the following day is deleted on loading as a probable data error.

For the worked example, this means: the price rises by at most 40.0% and falls by at most 42.9% in a day, and therefore does not trigger a warning. A real one-day pump that reverses the next day, on the other hand, would be silently removed. The check protects the calculation from data errors, not you from fraud.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.