I'd give this book 4 stars judging it on its merits as a book, but 5 judging it on how likely I think it is to improve your investing outcomes.
Taking those two aspects in reverse:
Smith's record as an investment manager speaks for itself: he's consistently outperformed the MSCI world index over the past 10 years (and the FTSE 100 has massively lagged the World, btw). He's also been way up there, if not top, in the ranking of funds like his own.
Perhaps this could have been due to luck on his part - perhaps there's some selective hindsight in reading a book by the best performer of the past 10 years, bearing in mind that old illustration of chance (I think it's in Benjamin Graham as well as elsewhere) about how if you get a load of people to flip a coin and then find the last person standing who (by chance) flipped nothing but heads every time, that doesn't make that person any better at flipping heads than everybody else - it's just selective hindsight. But Smith has not only delivered truly outstanding returns for 10 years straight, he's also convincing in setting out how he's done it.
The defining features of Smith's investing style and the main subjects of this book are his focus on "quality" and his disdain for, in particular, "value investing" (in companies with a low price to earnings ratio or price to asset ratio) but also the false dichotomy between "value" and "growth" investing in most media coverage of investing.
As you've probably read elsewhere if you're considering reading this book, the danger with value investing is that there are generally good reasons why companies are "cheap" in the sense of having low price to earnings or asset ratios: they may well be "value traps" whose share price never rises again or rises only after any investor has foregone a lot of gains through the opportunity cost of not being invested elsewhere. Such companies aren't in fact "cheap": they're expensive considering their poor return.
But growth investing isn't all fine either, if it encompasses investing detached from the fundamentals of the companies (such as "momentum" investing in anything with a rising share price), or takes no account AT ALL of the price to earnings and asset ratios, etc.
"Quality" companies, for Smith, are those that achieve high returns on their investments and have a good shot at growing their revenue and profit. This is a subset of "growth" investing - underpinned by a focus on the financials, the management, and the opportunities. These companies have to be available at a reasonable price to be worth investing in - although Smith sets out why a price-to-earnings ratio of say 20-25 in a quality company is likely to offer much better returns over the long run than a P/E of say 10 in a weaker company.
Which, again, you might already have picked up from investing media. What Smith does here much more than anyone else I've read over the past year or so of being interested in this stuff is focus very strongly on how much a company gets back from every dollar or pound it invests - its return on capital employed. He's much more focused on this than the P/E ratio.
That is, revenue growth isn't necessarily great if it comes at the expense of shrinking returns on capital employed. If a company is growing its revenue and profit, but getting less revenue/profit for each dollar invested (e.g. by acquiring other companies with weaker returns on capital than it has), it probably can't keep growing for very long, and anyway will offer worse returns than a company that can reinvest at high returns.
While I was familiar with the concept that it's fine for a growing company to reinvest earnings rather than pay a dividend, what was new to me was Smith's explanation of WHY this is better, and how much better it is.
Partly it's that dividends might be subject to taxes when either received or reinvested. More interestingly and I think importantly, Smith says, it's because when a company invests a dollar of its earnings in growing, it does so at the "book value" of the company. Whereas when you buy a company's stock, obviously you're paying whatever amount ON TOP of book value the market currently thinks the stock is worth. (Partly for this reason, he's generally not a fan of share buybacks - he thinks they're only worth doing if the company cannot do anything better with its money. But if it can't, maybe you shouldn't be invested in it.)
Put it this way: if a company is trading at 1 dollar a share in 2019 and can make 20 cents on every dollar it uses in 2019, but by 2020 it's trading for 2 dollars a share because the market has spotted what a great company it is, wouldn't you rather it kept that 20 cents and made another 20% on it in 2020, rather than giving it to you so you can reinvest it in the company at twice the price that cost you a year earlier? This cutting through the above-"book value" price of the shares also applies even if the share price hasn't increased. Essentially: you have to pay the above-book price when you first invest, but every reinvestment the company does after that is using your investment (well, your share of the company's capital) at book value. Even better, that translates into above-book froth/gains that you get to capture if you ever decide to sell. (Look, he explains it better than I do, although I did still somewhat have to take his word for it.)
So that's the five stars for the strategy.
Smith also has a clear, no-bullshit outlook and writing style. Why the book gets only four stars from me instead of five is that it's quite repetitive. It's a collection of Smith's annual reports to investors in his fund, alongside the newspaper articles he has written for the Financial Times etc over the past 10 years or so - the period of his fund's duration. Two of the articles were almost word-for-word identical; some paragraphs and quotes were reused multiple times in various articles; and the parts of the annual reports explaining the fund's costs were repeated in report after report. I see no reason why these repetitions couldn't have been edited out, unless the thinking was that people might read only certain chapters. But the bigger chunks of repetition in the reports are easy to skip, so it's not much of a hassle - and far outweighed by the book's qualities.