Direct Connection is the fundamental change in computing from the 20th to 21st century. In the 20th century, computers existed in a purely digital world, and humans acted as the input and output devices, translating the analog world into digital and feeding it to the computer, and then taking its digital output and interpreting it to act on the real, analog world. In the past 15 years, computers have begun to sense and interact directly with the analog world, without relying on humans for input and output, accelerating processes by millions of times. This has made thousands of times more applications viable, leading to the accelerating invasion of our personal and business lives by technology.

Autonomous vehicles. Imagine a car fitted with superb artificial intelligence, but where a human has to sit in the driving seat typing onto a keyboard what is visible ahead, while also reading instructions from a screen as to how to steer, accelerate and brake. The car would never leave the drive. It is not until the computer is connected to sensors and to the vehicle’s control systems that AVs become potentially viable and it is worth developing the AI to direct them.

Information Technology is being transformed from an industry vertical (like Healthcare, Banking, Retail, etc.) into an economic horizontal. In every other sector, increasingly the winners are those that behave like tech companies, investing as much as they can in technology to get ahead. This is sucking the growth out of other sectors, and concentrating it in IT.

We don’t need to. As an example, “fintech” companies are all investing as much as they can in technology in order to give themselves an edge in the Financial sector. Given this massive over-investment, it is unlikely that any of them are creating much shareholder value, but their huge spend is undoubtedly benefiting the tech sector, and we hold companies that are creating value for their investors as a result.

Again, we don’t need to. Even before a new theme emerges, the businesses involved will have been spending heavily on connecting computers directly to the real world, benefiting my Direct Connection stocks.

Technology has been outperforming the broader market for 10 years, and Direct Connection is such a huge change that it will go on doing so for at least as long again.

We look for four things: 1. A strong underlying trend that benefits the company, often Direct Connection 2. The ability of the company to create value from that trend 3. Barriers to entry 4. A reasonable distribution of that created value between stakeholders. Unlike other sectors, most technology investors focus entirely on 1 and 3, and ignore 2 and 4. This is a big mistake, as over a period of years 2 and 4 make an enormous difference to how stocks perform. It is therefore vital to look at value creation and distribution, and not just how fast a company is growing and how competitive are its products.

Software-as-a-service or cloud software companies: they grow very fast, as the customers get huge value from the cloud business model; but all the value that is left goes to the employees, because the market for talented engineers is so competitive. This is normally in the form of stock options (as investors seem to attach no value to them), and these constantly dilute investors’ ownership. The scale of this dilution is extraordinary: multiple years of free cash flow would be required to reverse a single year’s stock dilution; and the closer to San Francisco a company is based, the worse this dilution is – the correlation is quite striking!

We are looking for a strong underlying trend that a company can create value from, and a fair distribution of that created value between the various stakeholders including outside shareholders = our investors.

They are generally extremely expensive and often have very fragile business models. They tend not to create much value for external shareholders, and they are very dependent on continuous improvement in newsflow. If the news doesn’t go on getting ever better, they have a long way to fall. They are trades on short-term newsflow, not long-term investments.

Other technology investors would think our portfolio contained mostly rather boring names. But we are looking for businesses that are benefiting in particular from Direct Connection, where the value they create is very stable and the barriers to entry very high. These stocks will generally be quite expensive, but not expensive enough, as you can hold them year after year, steadily outperforming but without increasing their valuation metrics. We like to find stocks that we can buy and hold for 3-5 years.

We are generally fairly sceptical: companies with a market cap of more than $100bn have had an amazing 10 years; in other sectors that might be a good indicator that they will lead the field for another 10 years, but in technology that is not the case, as it is unusual for a tech company to stay at the top of the market for 20 years in a row. If you buy a basket of the megacaps, you own last years’ disrupters and collectively they will be losing out to next year’s disrupters. We therefore start with no megacaps when building a portfolio, and then pick the 3 or 4 that we have high conviction will do EVEN BETTER in the next 5 years than in the last 10 – and that is a very high bar indeed. So we tend to be significantly underweight the largest companies, and focus instead on stocks with a market cap of $1bn to $100bn, where the scope for upside is much greater if you get it right.