We're letting a good crisis go to waste. Since the financial crisis of 2008 there have been growing calls for financial reform, including calls to limit bankers' bonus culture, enforce tougher banking regulations and decrease risk-taking in the financial sector. This is a natural reaction. But these small changes are not enough.
One avenue for potential reform that has been completely ignored so far is to change the rules of the stock market itself. Stock markets have existed since the 12th century in France, and very little has changed in the intervening centuries. Essentially all new technology in the market has just improved transaction processing times. Do we really need a market that settles trades in milliseconds? With current technology, we can do more than make faster trades; we now have the ability to create better markets which have far less volatility and are just as liquid.
People are angry today. Angry at the banks and at the systems that plunged us into crisis. While it's tempting to condemn the entire financial system, the stock market is useful to society for two reasons. Firstly, the market allows promising companies to raise money for their plans through the issuing of new stock, and secondly, the market allows people the opportunity to invest in companies and receive a healthy return for making wise investments.
Profit from investing on the stock market can be broken down into two components: identifying value and identifying trends. When investors buy shares that are under-priced, those shares will either appreciate to a fair value, or they will pay a dividend that is relatively larger than other shares. This is the good side of trading. Investors can go to great lengths to understand a company's products and position in the market to better understand what a fair value for that company is. Society is better off when markets are dominated by value investors as money is channeled through to companies who are best able to use it.
The second component of profit from trading comes identifying trends. If you see that a share is doing really well today, and you buy that share, in the hopes that it will continue to rise then you are trying to identify trends. However, the key thing to note is this: trading on trends can make a profit, but it does not send money to companies in a careful, considered manner. When lots of traders try to identify trends then share prices jump around like crazy. Also, trends are a zero-sum game; here traders only make money if someone else loses money. This is the bad side of trading, as society does not benefit, wealth is not created merely sloshed around. I don't want to sound too moral here. I've traded shares based on trends, without an idea of the underlying value. It's just so much easier than trying to accurately value a large multinational company! However society is poorer off when bright people spend their time trying to identify trends in the market, rather than trying to meet society's needs.
Typically, trend traders only hold shares for a short while. Why hold it if there's another great opportunity to ride a share price tomorrow? With this in mind, many people support a Tobin tax, or a "Robin Hood" tax. This tax adds a small cost to each trade. The cost is practically non-existent for investors who hold shares for a long time, but it will add up very quickly for traders who want to trade daily. While this sounds like an ideal reform for the market, we need to beware of unintended consequences. The two most likely outcomes are reduced liquidity or increased volatility. Liquidity refers to how easily an asset can be sold, and if we impose a tax on people who buy and sell shares then it could very easily make it harder to buy and sell shares! The volatility of a share refers to how much the price fluctuates; if a trader knows that he must pay a tax, then he might wait for the price to drop lower before buying and he might wait for the price to rise higher before selling. Neither effect is good and in the current market structure it is very likely that one or both of these effects will occur.
This brings me back to the beginning of the article. Markets have existed in essentially the same form for centuries now. While the digital age has revolutionised many fields, what about the stock market? Stock markets now have many different options,futures contracts, credit default swaps, and other forms of financial engineering but everything is bought and sold in the same market structure. Arguably the only change that technology has brought to the market structure is speed.
Want to sell a share? Chances are the trade can clear in milliseconds, but not much else has changed. I propose that we now have the technology to fundamentally improve our market structure. We can now ensure that share prices are boring on boring days, while still allowing investors to buy and sell large numbers of shares without high price fluctuation.
Let's change the game by requiring investors to tell us how many shares they'd like to own for every possible price. Let's call this a "demand function." For example:
|Price||Demand for shares of company A|
Now, rather than an exchange receiving "sell" or "buy" orders from traders, the exchange receives updated demand functions. These demand functions could be simple, such as: "hold 100 shares if the price falls below $5, and hold 0 shares if the price is above $5." In fact, the simplest demand function could be: "hold 100 shares regardless of price."
Trend traders will want to buy more shares of a company as the price increases. I have already given reasons why we don't want people trading on trends, so we can start engineering the game by forbidding trend-trading. If you want to trade on this new market you have to be a logical value investor and as the price increases you have to demand fewer shares.
Given an exchange with many demand functions, how do we determine price? Well, each company has issued a fixed number of shares. We can add up all the demand functions and find the price at which the number of shares demanded equals the number of shares issued. This isn't a new idea. In fact, matching supply and demand is taught in every introductory micro-economic course, but now we're building that idea directly into our exchange. When the price of a share is known, then we can allocate shares to investors as specified by their demand function. If we were investing according to our example demand function, and the price was found to be 60¢, then we would be the proud owners of 40 hypothetical shares of company A.
The only constant in life is change. How should the price be adjusted when a trader changes their demand function? Well we can follow our procedure to determine what the new price should be. It is interesting to consider who holds how many shares now. For pedagogy, let us assume that a trader was positive about company A's potential, and for each price he has increased the number of shares he'd like to hold. This will push the price up just enough for others to demand less shares and the market will clear again. If the new price of our hypothetical share is 70¢, then we demand zero shares (as per our demand function) and we will sell all our shares at the higher price. This other trader will be required to purchase the extra shares at this higher price, and this money will be distributed to the traders who sold some of their shares. However, the entire market clearing process happened without any human intervention. This means that liquidity is built into the market.
"Not so fast!" you say. "What's to stop an undercover trend trader repeatedly sending new demand functions which buy more shares as the price increases?" Well, we can apply a Tobin tax to those pesky trend traders. In fact, the Tobin tax for a market of this type can be as large as 1%. We can specify that the tax only applies if your demand function has been updated in the last 3 months. This provides a strong incentive to consider the medium term when thinking about one's demand function, which is precisely the sort of behaviour we want to encourage!
Another great thing about this market model is that if the price of a share moves from $10 to $15 and then back to $10, then everyone who has left their demand functions unchanged will have made money. They will have sold their shares at the higher price and then bought the shares back at the lower price.
Shares also have value because they pay dividends. Currently the market pays dividends based on who owns shares at a particular instant in time. However a payout that is based on a single instant makes the value of a share far more variable in time. Our proposed market structure is designed to capture the slowly-changing value of a share. It is far more natural for the dividend payout to be earned in a similar manner to interest on a bank account. The amount you receive is based on how long you have held the share.
The sum of all the demand functions can also be interesting information. It could be used by the management of a company to better understand the market's belief about the value of the company. This could then inform decisions at the company about the issuing of new shares.
I suspect there are many more variations of a market that one could consider. Given how large an influence the market has on our daily lives, I'm surprised that there hasn't been more research into alternative market designs. Instead, market criticisms tend to fall into two camps: either the criticism is incredibly specific and doesn't see the forest for the trees, or it is far too broad and condemns the entire financial industry without providing feasible alternatives. Instead let's start a mature discussion of how the system can be improved.