Eventually, you will arrive at the stage where you want to start thinking about how your strategy will work in practice. This is the portfolio construction stage, he stage where you have to think about using your combined alpha vector to generate and update an actual portfolio. At every time iteration, your strategy is going to take an existing portfolio, make predictions about market behavior using your combined alpha vector and use this information to make trades that move the existing portfolio to an updated portfolio. There are more questions that must be answered when turning your market hypothesis into traits such as, how does your strategy take into account in various forms of risk? What quantities should your portfolio seek to optimize? Are there any additional constraints on your portfolio? How do real-world constraints like transaction costs eat into the theoretical return and how can these be mitigated? Let’s talk in a bit more detail about some of these questions. You probably already have an intuitive sense of what a risk is. It’s uncertainty about the future, and in particular, the possibility that something bad will happen, like losing the money you invested. In finance, risk usually refers to uncertainty or variability in returns and there are many different ways to quantify it mathematically. There are different types of risks. Risks inherent in the entire market are called systematic risks. One form of systematic risk is risk inherent to individual sectors, like the technology sector or the energy sector called sector-specific risk. For example, the success of the energy sector as a whole might be affected by fluctuations in the price of oil, but it’s unlikely that this would affect the technology sector nearly as much. Risk inherent to individual stocks is called idiosyncratic or specific risk. For example, a single oil company might be uniquely affected by political events in a region of the world in which it is exploring for oil. Your strategy will likely incorporate a mathematical model in these risks, which will help you understand and manage the different types of risks and attempt to avoid being overly exposed to anyone. Different investment funds have different mandates driven by the goals of their investors. For example, some funds will want to be neutral to all systematic risk. Whereas, other funds have a specific target for that risk. Your portfolio could be designed to maximize different objectives. As one example, you could seek to maximize expected return and minimize return variance. In this case, if two portfolios have the same expected return, you would choose the one with lower variance. Finally, your portfolio may be subject to additional requirements or constraints. For example, the policies of your investment firm may dictate that you can’t hold stocks below a certain level of market capitalization, or you may only be allowed to enter long positions due to government regulations in certain markets, or there may be a limit on what percentage of your portfolio can be invested in any single stock. These must be taken into account during portfolio design. Finally, in the training stage, you make actual trades in the market. After portfolio construction, you’ll have an ideal portfolio and a list of trades to make. In this stage, you’ll have to figure out mechanics like how fast to trade and to which time horizon you believe your alpha applies. You may decide to trade quickly and aggressively, or more slowly and passively. You’ll also have to take into account the fact that the trades themselves incur costs which will eat into your returns. This is the realm of market microstructure and high-frequency trading. Finally, spends market conditions don’t allow you to place the exact trades you intend. As you can see, there’s a lot to think about during this process. So, let’s get on with it.