STOCK PRICE MANIPULATION

is as old as Wall Street. In the mid-19th century, renowned robber baron

came up with a complicated scheme to profit from Erie Railroad shares that involved shorting the stock while at the same time greenlighting the issuance of Erie bonds. His efforts were facilitated by his position on the railroad's board. Gould went on to gain more infamy for trying to corner the gold market.

Times may have changed, but stock scammers are still driven to make a quick buck through any illicit means necessary. Thanks to technology, all it takes today to manipulate prices is a computer and high bandwidth. By sending out loads of spam using come-ons like "ride the bull" and "fast money" to hype tiny over-the-counter stocks, scammers drive up prices and then dump their shares. Subsequent buyers are left to discover the company's prospects weren't so bright after all.

Yes, it's illegal. Yes, regulators are cracking down on the practice. In March, the Securities and Exchange Commission suspended trading for 35 companies that were the subject of spam email campaigns. The move was part of a larger effort called "Operation Spamalot," in which the SEC is targeting microcap stocks traded on the Pink Sheets.

Despite the regulatory crackdown, the bottom line is stock spam works. According to research by Rainer B hme and Thorsten Holz in their 2006 paper, "The Effect of Stock Spam on Financial Markets," email campaigns are successful in temporarily boosting share prices and trading volumes. The day a stock spam email hit in-boxes, B hme observed a 215% increase in trading volume. And over the first few days from when the spam is received, the stock sees a "cumulative abnormal positive return," he adds.

Stock spamming is a profitable business model. The price run-up, however, lasts for only a few days, says B hme, an assistant professor at Technische Universit t Dresden in Germany. By the fourth day after the spam was sent, the price turns negative, says B hme, who is currently working to expand his research to include more data.

So who actually buys the stuff touted in spam? The study's authors attribute the dynamics in traded volume and prices to three groups: the spammers themselves, who trade the stock to profit from their campaigns; "naive recipients" suckers, essentially who believe in the faux investment advice; and "smart recipients," who try to get in on the action they know is coming from the other two groups.

"It is not irrational to buy the stock if you know the campaign is just about to start and you suspect the abnormal returns will increase in the next one or two days," says B hme.

We asked B hme about his findings, why spam works only up to a point and why it's not going away anytime soon.

SmartMoney.com: Your results show that stocks experience positive abnormal returns for the day of spam and two days after, then negative abnormal returns for the first, third and fourth day after the day of the spam. The movement seems a bit erratic.

Rainer B hme: We have a bit of fluctuation. It's puzzling when you look at the daily abnormal returns. We don't put too much emphasis on those daily returns. We work with the end of the data; it's much more interesting to look at the cumulative abnormal returns. From the cumulative point of view, the returns are positive until the fourth day. They turn negative on the fourth day. So it's around plus-two percentage points on the first day (the day of the spam message), then the fourth turns negative.

SM: Are the spammers the ones who are doing the buying and selling of the stocks?

RB: Our story is that somebody wanted to dump a lot of stock of these companies and they hired someone to promote the stock, to first increase the price and increase the volume. The lower the volume, the more market impact you have [on the stock price].... We think people who want to sell these stocks ask some agent to provide the volume, and these spammers create the volume.... The person who wants to sell the stock could be a shareholder, an investor or even the spammer. We don't know; we just observe the reactions.

So if you're a shareholder in a company and you want to get rid of stock, you can't sell shares without affecting the price negatively. So you want a stable price while you disinvest. These stocks are very illiquid. There are days when the company is not traded at all.... This is one of the main motivations: People want to disinvest, they want to create volume to allow them to sell at stable or inflated prices.

SM: What effect did the spam emails have on the volume of these microcap stocks once they were received?

RB: There's a clear reaction in both volume and return. We do two evaluations to measure volume. We take all spam events [emails] and see that volume [per stock] increases 215% on spam days.... We also take a more conservative measure. When we [excluded] spam messages sent after the markets opened (because it could be that spammers decided to spam based on what they already observed in the markets), we still report a 150% increase in volume. So spam days see higher volume than nonspam days.

SM: How do you account for why the returns turn negative after the fourth day?

RB: It is very difficult to statistically prove the significance of the change in price the more time passes from the spam day. Joshua Cyr tracks the long-term effect of spam in a virtual portfolio, where he bought every stock promoted in every stock spam email he got. He found that he would have lost 92% of his portfolio in the long run. He started in the beginning of 2005, and I think stopped adding new stocks in August 2006.

SM: You essentially prove that the business model behind stock spam works in the short term. But does the more spam sent equal more volume in the stock traded?

RB: We did a breakdown with the number of messages sent. We found that the more spam you send the stronger the reaction is. But we also find that the marginal benefit decreases, meaning the additional return for each extra message sent is less over time. When I've already sent 10 million messages, then the 11 millionth does not have the same effect as the first ones.

SM: Isn't it surprising that stock spam has much of an effect at all? Most people don't pay attention to those emails.

RB: We were not sure if stock spam was a phenomenon where people learn and it's fading out. We measured the reaction in the first half of the sample the first six months of 2005 and the second half. We found that the effect of spam is stronger in the second half. So when we wrote this paper it was still in an upward trend. That wasn't so clear in 2005. Now we know that the volume of stock spam was even higher in 2006. We were looking at stock spam at an early stage. Afterwards in 2007, it moved to other markets. At the time when we wrote the paper it was not clear if stock spam had reached its peak. We expect more results once our larger data set is available; we're including up to the end of June 2007. It's technically more demanding because spammers are getting much more sophisticated.

What's interesting is there's a constant cat-and-mouse game between filters and spammers. The filter developers try to catch all kinds of spam. The spammers try to adapt their messages in a way to evade the filters. What we see in the different technologies used by spammers like the ones with the images meant to distort the message is that the attempts to avoid filters are first seen in the stock spam. We think it's because stock spam is the most profitable. They're the early adopters of all new technologies.

SM: In the conclusion of your paper you attribute the dynamics in the stocks' trade volume and prices three groups of individuals: spammers, naive recipients, and "smart recipients," who try to participate in the price hikes. The smart ones are essentially trying to profit off the naive ones, right?

RB: It's an interesting question from a behavioral economics point of view. It is not irrational to buy the stock if you know the campaign is just about to start and you suspect the abnormal returns will increase in the next one or two days. It's unclear whether it's illegal or not; you're basically just exploiting public information if you act on a spam email. The question is whether the reception of spam messages is public or not. This is exactly how this herding behavior starts. But this works up to a point.... When the first ones start selling, it affects the price negatively, then causes prices to drop further.

We certainly do not recommend your readers to buy these stocks. You're always behind, you don't have real-time information. So for the private investor, it's hard to make money. You also don't know if you're in the first wave of the spam campaign or in the middle.... So it's dangerous to get in this game.