The quality of underlying data is very important when you try to make decisions based on that data. For instance, my friend Karthik Shashidhar writes at Pragati that wage rates in rural India vary so much that one tends to disbelieve the underlying data itself.
I’ve often maintained that price data supercedes most others – that is, if you know that a price was traded, it’s often a better indicator of what people value something today, than any theoretical measure.
For instance, you could theoretically calculate the long term growth rate of, say, a company making a toll bridge using characteristics like traffic estimates, bridge maintenance costs, toll increase rates and so on, and come up to a figure of profit they will make in the future, and “discount” that profit to the current day giving you a per-share value of the company. And then, the price in the market could be less than half that. The real indicator of what people value the company is the market price; anything in theory will be riddled with assumptions and fallacies.
But here’s the crux. Even the price, sometimes, is unreliable.
If there’s too little volume, price is unreliable. Obviously, if a stock trades just Rs. 50,000 worth of shares a day, I might be able to manipulate the stock by placing an order of just Rs. 10,000 – and why would I want to do that? Because, say, I’m a mutual fund and a higher price will enable me to report a higher “NAV” (Net Asset Value) which investors use to compare mutual funds. As long as investors don’t leave in droves, I’m safe. (And if they do, I’ll blame the investors)
Another price unreliability is when you take “quoted” prices, not real ones. This involves asking people what they would pay for something and then assuming that’s the price.
You think this is crazy? But it’s how LIBOR works – some people call up some other people what they would pay for an overnight dollar loan (and other stuff), average the answers and assume that’s the real price. And hundreds of thousands of contracts around the world fix their interest rates to that number.
In India, the USD-INR (Dollar-rupee) contract on the NSE is settled at expiry with the RBI Reference Rate. The RBI Reference Rate is an average of the “quotes” provided by banks to the RBI which calls them between 11 am and 12 noon. The Banks themselves are traders on the NSE, and will have open positions – given that this is a quote only (no buy or sell will happen) will the banks not be incentivised to tweak the quotes, even marginally, to benefit their positions?
A third factor of unreliability is when you use averages to smooth out volatility. What you hear about as the price of a security on TV after market hours, is the closing price. This is not the last price traded. This is the “volume weighted average price of the last half hour” of trading. See my video on VWAP:
When you use a VWAP price as the close, the close price is not actually the price of the security. Combined with low volume, it is unreliable in itself. Also, if you have ONE big trade at 3:01, you might see the VWAP closer to that price rather than the price that it moved to in the remaining 29 minutes.
You also have market manipulation. You could get two brokers to trade between each other, in a thinly traded stock, and keep increasing its price. This attracts pure-price watchers, and when they dive in, the operators exit, leaving the price-trader holding a dud stock. There’s no easy way to identify such stocks, since it happens in nearly all stocks. As a rule of thumb, I wouldn’t use only price to buy into a fast-rising low-volume stock.
However many the disadvantages, the unreliability of other data is worse. Can you believe “financial statements” made by management instead? Everyone lies, even auditors are in on it, and there is no punishment of any sort. Will you rely on “official estimates” by big brokers or banks instead? Again, they’re out to bullshit you because you, my friend, are the sucker in the game.
Marrying the price doesn’t mean you’re not wary of its infidelities.
