Digital Business Transformation Gone Wrong

Last week, my colleague Patrik Löwendahl gave a brilliant presentation on the future of digital business at our annual Digital Tomorrow event. I think that there is a lesson to be learned by other thought leaders from his speech. Without mentioning the buzzword “The Internet of things”, he gave several examples of applications that build on the ubiquity of smart devices that will sense and interact with us and each other. I have to admit that I have heard the same story from Gartner on a recent Customer Strategy event. But, contrary to Gartner, Patrik challenged the audience with a thought-provoking question after each demonstration. His question, “Cool or Creepy”, should be asked more often. In this blog post, I will argue that we tend to be naïve when pushing for digital transformation. We tend to think that the digital business transformation always is driven by the competitive pressure to retain and win customers. Consequently, it is assumed that customers always benefit from the digitalization. In my opinion, that is not always the case and I am concerned that we embrace digital advances without looking at the consequences. I will prove my point by putting the spotlight on the most obscure, most advanced and most profitable winners from the digitalization of the capital markets; the High-Frequency Traders (HFT). I will share the insights that my favorite author Michael Lewis presented in his most recent book, Flashboys, to illustrate the risk that the digital business transformation goes awry.

I guess that everyone that has studied micro economics has read the same article on the economic organization of a Prisoner of War camp during the Second World War. It illustrates how the prisoners fast forwarded our economic history and quickly established an effective market where cigarettes where the common currency and the price for goods where quoted on boards. Similarly, it didn’t take long time until some prisoners started exploiting arbitrage opportunities. They realized that they could buy canned pork cheaply from Jews and then resell them to others at a profit. Though despised by the other inmates, the arbitrageurs filled a purpose in bridging the gap between different markets. In my naivety, I assumed that the High-Frequency Traders (HFT) would fill the same function as the POW arbitrageurs when they emerged in year 1999 as the US Securities and Exchange Commission (SEC) authorized electronic exchanges. When working as a consultant at the leading Investment Bank at the time, I witnessed the explosion of new electronic trading places, Multilateral Trading Facilities (MTF), e.g. Burgundy and Chi-X, at the same time as the banks created their own internal markets that usually are referred to as Dark Pools. Many of us feared that the many trading places would dilute order volumes and increase the spread between the best buy and sell orders and thereby make it harder to provide investors with the best possible price. But, looking at the trading statistics, you get the impression that the HFTs have used their lightning fast computers and connections to create a single virtual stock exchange where the best bid and ask prices are mirrored immediately. Michael Lewis tells another story in Flashboy about some traders at Royal Bank of Canada that started to investigate the flaky market behavior they experienced as soon as they entered orders in their Order Routing systems.

They proved that the HFTs baited the markets by flooding them with small buy and sell orders so that they could intercept any genuine interest to buy or sell stocks. Let me illustrate with an example. An investor asks his broker to buy 10.000 shares in Microsoft at market. The best Ask price is currently $ 40,21. The broker sends the order through his order routing system to the market that is the most beneficial to the bank. The buy order will be matched with the lowest sell-price ($ 40,21) but will only be partially filled with 100 shares as the HFT quickly cancel his remaining sell orders. Next, the HFT will race to the other markets to buy the remaining 9.900 stocks ahead of the investment bank. The race is about milliseconds but the HFT is predestined to win thanks to faster algorithms, dedicated lines and servers that are located in the same data center as the stock exchange. So, when the unfilled investor order finally reaches the markets, the HFT is waiting to sell the investor the 9.900 stocks at higher price, say $ 40,25, and pocket a $ 4K profit without taking any risk. Consequently, the investor has been forced to pay $ 4K more than needed if the HFT had not interfered with what used to be a well-functional market for genuine buyers and sellers.

This electronic front-running does not only deceive the investors. It creates an illusion of a highly efficient market with liquidity and small spreads when all trades are duplicated, orders are entered to lure investors and the thin spread is only an effect of HFTs that want to stand first in line to deceit genuine investors. You really start to question the effectiveness of the markets when you hear that the HFT firms accounted for more than 99% of all orders on the US stock markets but only 50% of the trades.

To me, the scariest part in the story is that all parties on the stock market knew that the investors were getting screwed but did not act, on the contrary, they tried to get a piece of the cake. According to Michael Lewis, the investments banks got their piece by selling access to their dark pools to the HTFs, the exchanges offered co-location and introduced weird order types to win HTF transactions and even the SEC officials were in bed with the HFT firms. The unhealthy linkage between the HFT firms and the regulators at SEC become obvious when the team from Royal Bank of Canada presented their solution to prevent the front-running. They had programmed an order router so that it delayed orders to markets with low latency to ensure that split orders reached all markets simultaneously. The reactions from the officials stunned the team. The officials disapproved and argued that it was unfair to both HFTs and investors that orders were delayed. To them, it was all about speed, regardless of who won the race. Their reactions raise many questions but I settle for one: Cool or Creepy!

Categories English, UncategorizedTags ,

4 thoughts on “Digital Business Transformation Gone Wrong

  1. Interesting, but to be honest a somewhat misleading picture of the state of the markets. It’s just silly to compare the cost of trading to some ideal state where spreads are zero and order depths are infinite. Instead, we should compare to a situation where HFT isn’t there at all. Let’s call that situation “1999”, because at that time the HFTs were in fact not there.

    Was it cheaper to trade then? Hell no, it was many, many times more expensive! The spreads were higher and “specialists” were making an absolute killing. Surely no one would want those days back? (Or at least no one except the dinosaurs, a.k.a. the specialists.)

    The problem is, as always in the financial markets, that it’s OPM (other people’s money). So you are an asset manager and you steal 2% of your clients’ money each and every year. (They don’t like to call it “steal”, but they do take the money and they offer very close to nothing in return, so it’s a sadly accurate term.) They get their money from an excellent sales effort from sales people who are dishonest enough to not even call themselves sales people, instead opting for “advisors” and other completely misleading terms. In this situation, they simply don’t care about the 0.02% they give away to HFT.

    For me as a small retail investor, the current situation is fantastically much better than the old situation. The cost of trading has fallen dramatically.

    The relevant figures to compare should be the cost in three different scenarios:
    1. Today, where people managing OPM doesn’t care
    2. Today, where people managing OPM does indeed care about execution
    3. The situation 15 years ago

    If you compare those costs, you will of course find that 3 >> 1 > 2. So the main problem is that pepole managing OPM simply doesn’t give a s**t. Trying to blame that on HFT is sad in so many ways.

    Finally, the term “electronic front-running” is a weasel word that should simply not be used in an honest discussion. Front running is a well defined term for an illegal activity. HFT does not engage in that. Instead, they pay (dearly!) for information and use that to make educated guesses about order flows, but do note that it’s still guesses! Suggesting that this is illegal or even unethical is downright silly. Some people pay dearly to Bloomberg and Reuters for information they hope will give them an edge in predicting market movements. Is that also illegal or at least unethical? What is, exactly, the difference?

    Michael Lewis is my favorite author too, but “Flash Boys” was terrible. The lack of actual comparisons between costs under different regimes was completely unacceptable. We all know Mr Lewis can count, so the implication is that he avoided doing so because it would clearly reveal that the figures don’t point in the same direction as his narrative. That is a bad sign for his narrative.

  2. You are correct that front-running is illegal. It is easy to stack evidence that they actually are performing front-running. The test that the team at Royal Bank of Canada did where they delayed orders to some markets to ensure that they were delivered simultaneously to all markets is probably the most indiscriminating evidence to the HFTs. All of the sudden, they could actually close their orders on the bid and ask quotes displayed on the screen. And the fact that representatives from HFT firms brag about their track record and claim that they have not lost money on any single trading day since their inception should rule out any notions that they take any risks. True, the ruling should be innocent until proven guilty. But who’s to judge? The Flashboys made an investigation and identified more than 200 employees that have left SEC to join HFT firms. I doubt that anyone that considers working at a HFT firm would initiate an investigation to their trading strategies.
    No one came out clean in the book. I understand that you and others that are dependent on the markets places want to put the blame upstream. Especially considering that the HFT firms accounted for two thirds of the profits at some of the US based exchanges. The fact that the HFTs pay dearly for co-location and push for new freaky order types where the sole purpose is to poke the market further indiscriminate the HFT firms. Micke, you said it yourself. This is a question is on ethics and legislations not the cost comparison. At least not to me.

    1. I think we need to be quite a bit more strict in the definitions here. There is a world of difference between *knowing* that a customer is going to do X and *guessing* that he is. The first is illegal, the second isn’t. The first is what broker desks routinely do. The second is what HFT do.

      It is true that the RBC “experiment” demonstrated that this is how HFT act. However, what RBC could have done would be to place one big order on one exchange and then see that the HFT algos would have done exactly the same thing, i.e. pulled their orders on that level and offered new, slightly worse prices. Had they done that (which, btw, I am completely convinced they did, but Lewis “forgot” to mention that), everyone would have seen that they are in fact guessing.

      The fact that HFT makes trading profits every day doesn’t mean that they are not taking risks. It means that they take many small, calculated risks every day and trust the law of big numbers. The same goes for ICA, of course. Their revenues have been higher than their costs of good sold every single day for the last 50 years. Are they also front running? Is ICA not taking any risks in their business? Also, what sort of contrived accounting should make us compare trading profits without also considering fixed costs? The fact that HFT brag about something irrelevant is a little sad. The fact that other people doesn’t even understand why that is an irrelevent statement is very, very sad..

      I truly don’t see the difference between paying for information and acting based on guessing how that information will affect markets on the one hand and on the other hand, well, no there is no other hand. This is exactly what all market participants do! There is a difference in how quickly they act, and in what types of information they use, but not in what the essence of their business is. While manual broker desks may or may not have access to information that isn’t publically available, we know for absolute certain that HFT only have access to public information. You could buy that at the exact same conditions they do. Too expensive, you say? That’s what I think about a Bloomberg terminal. Ami right in wanting to ban you from having one, then?

      What annoys me about this is that we know that by far and away the worst part of the finance industry is the managing of OPM. They add no value, suck ridiculous amounts of money out of the economy and they fail to keep up with the times. And when they, like in this book, get called out on it, it is the people exploiting their laziness and stupidity who are somehow to blame!?

      Why would anyone in their right mind defend people who are so lazy as to not even bother figuring out how the market they operate in actually work? Who else but an OPM manager would get away with that?

      If this is a question about legislation, I think we all agree that the SEC is terrible. REG NMS is a disaster. Basel II was worse, and I can guarantee that Basel III will be even worse. But just because legislation is hard and terribly done doesn’t mean we should blame those who saty well within the confines of the laws. (There are many HFT firms. I bet it has happened that at least one of them broke some law at least once, and if so they should of course be punished. But I am also very certain that the prevalence of law breaking is orders of magnitude less common and less serious compared to at the big banks. LIBOR fixing – need I say more?)

      1. True, the HFTs are guessing. They face the small risk that the investor will cancel his unfilled order due to a tight order limit. The strict definition of yours is probably the excuse that many use to maintain status quo, a very lucrative status quo to everyone but the investors. IMO, ethics cannot be as narrowly defined. Let’s use your analogy with ICA. Let’s assume that they installed cameras in the shelves to capture information on your intent to buy their goods and they used that information to set the price on the shelf. The prices would be low as the customer scans the shelves but as soon as he lifts an item from the shelf he will notice that the price for that item increases. Is that ethical? If not, when does it get unethical? At the POS where the customer has the last chance to cancel?
        Farfetched? Probably, but the grocer could use the same arguments as the HFTs. The initial low price on the shelf was only to see what goods you are interested in. It was not a commitment to sell and he has the prerogative to use his lightning fast technology to scan your grocery list and buy the number of items you want ahead of you from the cheapest competitor to ensure that they go out of stock and that you have to accept his new higher price.
        I do not defend the investment managers. This blog post was on the High-Frequency Traders but I hope that you start your own blog and write about the management of OPM. I promise that I will be your first follower .

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this:
search previous next tag category expand menu location phone mail time cart zoom edit close