Stock Asset Returns Are Predictable Part 1

Theoretical Background Since the origin of Pharma (1965 and 1970), from the viewpoint of information enforcement, the efficient market is "all available information" (Fama 1970) and "stock price" which fully reflects the market efficiency classification It is defined as. There are 3 levels. Weak format, semi-strong format, strong format. First, the information set by weak form efficiency reflects past price or return only. Secondly, the information set in a somewhat more efficient format contains information available to all market participants.

Stock markets are not always in a bubble state. The shares give some ownership to the owner's large enterprise. A real company has a real benefit is that you have real worthworthy assets. In fact, if we can accurately predict the company's profit, we can accurately calculate the value of the stock. Future revenue is the reality that the stock market price is based on. However, investors may not agree on where the proceeds are going, so some think that the stock is cheap but others think the stock is expensive. As the news aired, some investors changed their mind, which led to fluctuations in stock prices.

Suppose the future price-earnings ratio is very small, like 0.01 (1%). There is a model to predict the stock price in the future, and profit and loss is directly related to the influence on the forecast. How can we measure the loss associated with the prediction of the model and the subsequent future projections? The squared error loss has nothing to do with the flag, prediction - 0.01 is the same as prediction 0.03. If you make a bet based on the model's forecast, the forecast 0.03 will give you a profit, but the prediction of -0.01 will lose money, but the loss will not be recorded. Considering the signs of prediction and actual values, better loss is necessary.