Recently a video appeared on cnbc.com with the headline, “Markets surge on ‘phase 1’ trade deal hopes” and is typical of news reports that appear almost daily in the financial press purporting to tell us why securities prices rise or fall. Their titles are worded in such a way as to suggest that the journalists who write, or present, them somehow know exactly what caused the price changes in the securities markets they are reporting about. But, of course, this is nonsense since in order to do so these reporters would have to read the minds, or in the case of artificial intelligence-directed trading know the algorithms, of the millions of traders active in the market on that day.
When reporters say the “market” rose or fell they are talking about the movement of a price index the composition of which is to a great extent arbitrary. But, even for an index that covers, for example, all equity securities, any changes of this composite index are a result of changes in the prices of all of the individual shares of which it is composed. And the reasons for the rise or fall of each of these component stocks are largely different from each other at any particular moment. There is no single reason for changes of market index levels or of the prices of individual assets.
And the “reasoning” used by artificial, or machine, intelligence (an important tool of trading firms nowadays) to decide whether to buy or sell, when, and at what price are not always explicable. For example, we cannot translate into words why a neural network makes a conclusion. The “opinion” that a neural network will form about data presented to it is subtly encoded in the strengths of the connections, or synapses, between its neurons. And these connection strengths will change as it is trained on new and different data. How the matrix of these connection strengths at any moment in time can be interpreted and transformed into verbal reasons is not known at this time.
The limits of knowledge and of introspection
We often cannot provide the reasons, or even one reason, for our own actions, let alone give the reasons for the actions of others. Almost all of our mental processes are unknown to, and inexplicable by, our conscious mind. This is because our brains are “natural” neural networks. The responses our brain will give to future stimuli, and its memories of previous stimuli and responses, are encoded in the strengths of the connections, or synapses, between its neurons. And these connection strengths are continually and rapidly changing as our brain functions. How to translate these connection strengths into verbal explanations of why our brain has reacted, or will react, in particular ways when presented with particular stimuli is unknown at this time. Thus we are unable to give so called “reasons” for our past, present, or future actions and certainly not for the actions of other people.
An artificial or natural neural network (i.e., a brain), cannot know or “see” all of the workings of itself as they are occurring at any particular time. It cannot be aware of the totality of the processes that comprise itself at the same time that these processes are occurring. Just as the eye cannot see itself directly, but only by reflection in a mirror, so too we (i.e., our brain) cannot have direct knowledge of our entire self (our brain) in real time. Therefore, since we cannot be aware of all of the events occurring in our brains, we also cannot possibly know for sure what the reasons (i.e., causes) are for our actions or even for our thoughts. But, although we cannot ever have direct knowledge of all of the events in our brain as they happen, we might conceivably become aware of the traces, or memories, of some of these past brain events or brain states (after they have been encoded and stored as synaptic strengths of the neurons of our brain) and thereby have what is called an introspection or awareness of self. However, this self of which we think ourselves to be aware is not the self existing at the moment of our supposed introspection, but is instead only a representation of our past self. And whatever memories we do have, or think we have, of our past selves will be very incomplete, faint, and indistinct.
So, if we cannot observe all of the processes of our own brains as they are occurring or even learn of them afterwards, and therefore cannot know all of the reasons or causes for our own actions or thoughts, then we most certainly cannot discover the reasons for the actions or thoughts of other people whose brain states are completely unknowable to us.
Buyers and sellers have differing reasons
And even if we could know the reasons why people act as they do we would soon discover that buyers and sellers in markets always have differing reasons for their actions. This must be the case, otherwise they would all want to be buyers or they would all want to be sellers. There is then no one reason or cause why the price of an asset, or of a market index, will go up or down. So, it makes no sense to say, as in the headline quoted above, that an overall price rise or fall is due to some single event or reason. It would be more accurate to say that price movements of assets, and of composite market indices, are the net result of bargaining between many people of differing opinions and tastes (which are constantly changing) and that we cannot have any direct knowledge of these opinions or tastes.
Again, price changes in markets are the result of the bargaining process between many buyers and sellers. If there are more market participants wanting to buy than to sell then the price of the asset in question may rise as a result of the bargaining process. This may also happen if the strength of the desire of buyers to buy or of sellers to sell changes. And, obviously, at any particular time buyers will necessarily have reasons for wanting to buy that are different from the reasons sellers have for wanting to sell.
So why do journalists give reasons at all?
It may be that business journalists attempt to give some simple or single reason, or any reasons at all, for market price rises or falls because their readers demand it. This does not mean that readers literally make their demands for these simple explanations heard by, for example, telephoning the news services. What is meant here is demand in the economic sense, such that publications which do provide simple explanations for events that have complex causes, like the price movements of assets or of price indices, will attract more readers or viewers, will thereby attract more advertising revenue, and so will be more likely to stay in business than news services that don’t. So, what arises is a sort of journalistic darwinism where only the news organizations with the simplest and most satisfying (although not always the truest) explanations will survive in the journalistic marketplace.
If the journalists know, then why aren’t they rich?
If business reporters knew why securities prices or indices rise or fall on any particular day, or how investors will react to news events, then these reporters could probably get rich very quickly and easily. The fact that they are not fabulously wealthy shows that they and their news organization don’t really know what the reasons are for market movements. And, if they did know something that could make them rich they certainly wouldn’t publish it to millions of readers or viewers like you.