Wednesday, November 2, 2016

Quantitative Momentum, by Wes Gray and Jack Vogel

This is a terrific book. As good as "DIY Financial Advisor," which was also great.

I did, however, manage to get myself all sorts of confused about which approach to measuring momentum the authors use in the book. I believe this would not have been a problem for me if two approaches to measurement hadn't been laid out early on, on page 11.

There, the authors start out saying they're going to outline what they mean by stock selection momentum. Then they define two approaches to measure momentum, time-series versus relative strength. The explanation is clear. Although the conclusion is puzzling to me, as on page 12 they say that these two approaches are often used in market-timing or asset-class selection, neither of which is the focus of the book.

But surely in this book they do measure momentum, and presumably with one or both of the two methods they just defined. So which is it? I was hoping for a summary line like this: "When we talk about measuring momentum in the rest of this book, we'll be using the {choose one of: time-series / relative strength}approach." Or, we'll use both and use words like "trend following when we use relative strength" and "generic when we use absolute," or whatever would be accurate.

I figured this would become obvious as I kept reading but I didn't come across a clear rule. Yet it must matter, since the authors bothered to define the terms. I find myself wondering, perhaps this is an OCD trigger issue for me. You're no doubt wondering the same.

On page 49, it sounds like relative strength is the approach. On page 77, the authors introduce the phrase "generic momentum" as a time-series approach. Since they earlier spent a page defining time-series and relative strength, I was really hoping they'd come back to those terms again. Now I have three terms, but I feel pretty confident that I can reduce them to two, with time-series and generic as roughly equivalent.  This was supported by "How to calculate generic momentum" on page 80. I'm not super confident though, because I kind of feel that if they wanted me to consider generic as isomorphic to absolute, they'd have said so back on pages 11/12.

Things sort of come together on page 122, where my sense is that the authors calculate generic momentum (time-series) and then use those scores to do relative strength measures against the universe of stocks. Well, whether or not that's what they meant, that's what I took out of it. Since I wasn't clear on the fundamentals from page 11 to page 122, I'm not confident that I have it right now either. And on page 172 they specifically mention time-series, but that may be to clarify the method used by the reference for that particular analysis.

In fairness, I am so much not the target audience of professional investment quants that I'm clearly not a good test subject for the readability of this minor detail within this book. But if the authors ever do another edition, in deference to the slower students, they really should consider changing the summary section on page 12 to something like this:

"You'll find that we also use the term 'generic momentum' as a synonym for time-series or absolute momentum. As you progress in this book, you'll find that we use a combination of the two (time-series and relative strength), using a time-series sort first, and then comparing the outcomes to the universe of stocks to get relative strength as the secondary measure."  [Unless this is dead wrong. Sigh.]
I'd also probably delete the sentences on page 12 that say these usually only matter to market-timing or asset-class selection which aren't the subjects of this book -- because it feels like the authors are telling the reader that they've wasted their time understanding a description of time-series and relative strength when it isn't even relevant to the book they are reading. And lead them to wonder what the authors do use.

Okay, barring that, I really loved the book!  Well written, very clear, and good examples.

Tuesday, November 1, 2016

First Strike, by Ben Coes

I didn't enjoy the last book I read from this author, so I'm not quite sure how I ended up with this one. Must have been an eye-catching display at the public library. My comments on the last book were, "...suffers the typical trite memes of the "special forces operator acts as spy to save the world" genre."

Well, not much has changed. The hero is a bit less stoic, due to a love interest. But the plot suffers from ridiculous hero worship (the US President is a personal fan) and absurd plot devices (our injured hero flies directly to NY City to save the day personally, presumably because there is no other special forces operator on the planet who can do the job). Oy.

Okay, I will remember Mr. Coes and try much harder to avoid his novels so as to avoid getting all worked up.