The idea is very simple and it’s certainly not anything new. (Location 54)
Tags: orange
This book is about systematic equity momentum. (Location 69)
Tags: orange
If you like to construct realistic simulations of your trading strategies before deploying them, which I really do recommend, momentum strategies are very complex to model. (Location 71)
Tags: orange
They trade according to similar principles that I outline in this book. (Location 86)
Tags: orange
The problem is that they don’t. Yes, you’ve got thousands and thousands of stocks to choose from, but when it matters they will all behave like reindeers. (Location 314)
In normal market conditions, stocks can appear reasonably independent. When we have a bull market, most stocks go up but good stocks go up much more. (Location 317)
In a bear market, this quite high correlation between stocks suddenly approach one at terminal velocity. (Location 320)
This destroys the very idea behind diversification. (Location 322)
The lack of diversification in equities is a factor that it’s critical to be aware of. (Location 327)
The more stocks you hold, the closer your strategy will resemble the index. (Location 328)
Of course the word ‘momentum’ is not part of the official Standard and Poor’s index methodology. (Location 335)
The reason that a stock is part of the index is that it had a strong price development in the past. When a stock leaves the index, it’s usually because it had poor price performance and dropped below the market cap requirement. (Location 339)
When you look at a long term chart of such an index, you’re looking at a momentum strategy. Strong performers are included, poor performers are not. (Location 342)
This all means that the indexes make the equity markets seem better than they really were. (Location 345)
The problem is of course that this stock wasn’t in the index then. It’s only in the index now because it had an amazing ten years. You probably wouldn’t even have heard of the stock ten years ago. (Location 349)
Most time series sources available to the general public ignore dividends. Odds are that most stock charts you’ve ever seen, and perhaps even all, completely disregard dividends. (Location 371)
You may wonder why we assume that dividends are reinvested. Well, because any way you cut it, you’ll have to make some assumption about what happens to that cash, and that assumption will inevitably end up to be wrong. (Location 407)
But it does matter to the type of long term momentum approach that this book is concerned with. There are two problems that come from disregarding these adjustments. (Location 424)
Picking an index is more important than it might seem. The index is your benchmark. It doesn’t mean that you have to track the index, but it’s a yardstick to measure your performance against. (Location 443)
A stock with a higher share price will have a greater impact on the index. This is rooted in the very old way of thinking, that the stock price itself has some sort of analytical implication. (Location 459)
All of these indexes are weighted based on market cap. The higher value a company has the larger weight it will get in the index. Whether or not that makes sense depends very much on your point of view. (Location 491)
Classifying stocks by sectors is just a way to keep track of what the companies actually do. It’s a good idea to be aware of sectors even if you don’t apply any fundamental analysis. (Location 506)
I tend to use the GICS scheme, since it’s a coherent global standard and easily available in most market data platforms. (Location 514)
The concept of trend following was originally developed for futures trading. (Location 523)
Trend following is conceptually very simple. When prices start moving in one direction, whether up or down, you jump on the bandwagon. (Location 525)
This dumb strategy has however shown very strong results for the past thirty odd years, and according to some research it’s even performed remarkably well for hundreds of years (Kaminski & Greyserman, 2014). (Location 534)
Classic trend following is done on futures. The normal way to execute this strategy is to follow trends on a large set of futures markets, covering all major asset classes. (Location 543)
Any given market or even asset class can, and often will, have extended periods of time where trend following doesn’t work. (Location 547)
The core premise of trend following is therefore based on diversification. (Location 549)
The big money in trend following is made on the extreme trends. (Location 556)
In the end, trend following boils down to statistics. It’s about making sure that you’ll have favorable probabilities of gains in the long run at acceptable volatility. (Location 560)
Futures trading allows for very high leverage. Even more than that. It allows you to completely disregard leverage. (Location 572)
Most professional traders only use 10-20 percent of the cash as margin. (Location 576)
The real killer is correlations. Stocks are very homogenous as a group. (Location 590)
Whether you hold ten stocks or fifty, you’re still mostly long beta. (Location 593)
The war stories about making money on buying the right stocks are told over and over again in the media and in the blogs. (Location 596)
On the short side, you’ll find the opposite. Your position will shrink for every day it moves in your favor. (Location 610)
The other reason for short positions being troublesome is more straight forward. They’re just not as well behaved. (Location 613)
Trading the short side is difficult in all asset classes, but most of all in stocks. (Location 622)
This trading model can go long and short. It goes long in a positive trend and only short in a negative trend. (Location 652)
Yes, it’s the short side. The short side of regular trend following, as applied on cross asset futures, is a concern. (Location 930)
Can we forget about the short side and move on? Good. For the rest of the book, there will be no more shorting. (Location 938)
Actually we don’t get more money than the index. I was just teasing you a bit, and trying to make a point. (Location 1060)
That means that a total return index will show higher return than a price index. During shorter time periods the difference may not be that large, but once you’re dealing with multiple years or decades, the difference will become quite substantial. (Location 1071)
This trading model shows pretty decent results. It’s still not a recommended approach and it still has lots of kinks to work out, but it’s better than applying a classic futures model. (Location 1106)
As could be expected, this model suffers losses during bear markets, such as 2000 to 2003. (Location 1110)
Trend following on single stocks is a bad idea. Your success will be completely luck dependent. (Location 1250)
Even for such great performers, most trend models didn’t work that well. Sure, the stock price might have gained massively in a decade or two but usually there’s heavy volatility along the way. (Location 1256)
Equity momentum strategies versus trend following. (Location 1287)
If you expand the scope of trend following to mean ‘buy anything that makes money’ then trend following is every single trading strategy you can think of. (Location 1289)
It’s about buying positive trend and shorting the negative trends, usually employing a trailing stop loss or exiting when the trend strength fails by other measurements. (Location 1291)
Usually, if you look closer, you’ll see that what they’re doing is a lot more like momentum strategies than trend following. (Location 1295)
When a stock has been going up for a while, the likelihood of it continuing up is greater than for it to turn around. (Location 1302)
Everyone likes a winner. (Location 1306)
Anyone who’s ever bought or sold a stock knows of course that this is not the case. Stocks move up and down all the time and it’s very difficult to explain at the time why they are moving. (Location 1308)
I’m certainly not dismissing the many excellent researchers and fundamental investors out there of course. (Location 1322)
The other issue is that becoming a skilled fundamental researcher tends to mean a high degree of specialization. (Location 1327)
Momentum investing is about buying what’s moving up. When the price is increasing, we buy in expectation of the share price continuing to increase. (Location 1333)
There has been plenty of research as to the causes of momentum returns. It’s not very difficult to show that the momentum effect works or at least has worked so far up to now. It’s trickier to explain why. (Location 1336)
returns. It’s not very difficult to show that the momentum effect works or at least has worked so far up to now. It’s trickier to explain why. (Location 1337)
There’s been plenty of research confirming the momentum effect, both from academics and practitioners and there’s not a shortage of well performing equity momentum products. (Location 1351)
A crucial point to keep in mind is that the momentum effect will in reality work very differently in a bear market. (Location 1355)
Buy the stocks that are moving up, and don’t buy the ones not moving up. The obvious problem is that perhaps the stocks you know are not the most interesting ones. (Location 1366)
There’s a substantial discretionary element involved in visually looking at charts. This has the potential of introducing a random element, which can swing either way. (Location 1372)
Earlier I mentioned that it’s not a good idea to hold momentum stocks during a bear market. (Location 1391)
If you’ve managed to come up with a plan that covers all these factors, then you have a real trading strategy. (Location 1394)
Almost all stocks are impacted on a daily basis by the overall state of the market. (Location 1409)
In bear markets, it suddenly doesn’t seem to matter which stocks you have anymore. (Location 1414)
If you intend to hold a momentum stock portfolio, or any other stock portfolio for that matter, you always need to be aware of the prevailing market regime. (Location 1420)
The critical point is that you need to have a long term market regime indicator. (Location 1430)
Practically all equity portfolio strategies can be improved significantly by simply adding this one rule. (Location 1435)
It’s a much more difficult one and certainly a riskier one. (Location 1447)
The moving average here will be used as an indicator of market regime. (Location 1450)
While the core scope of this book is about momentum, the principles can be used for other styles as well. (Location 1465)
Take a very common ranking method found on various internet websites. A popular method seems to be to rank by the percent difference between the price and a moving average. (Location 1470)
Volatility is very important. This game isn’t about who makes the highest absolute return in a year. It’s about who has the most return per unit of volatility. (Location 1477)
This leads to the obvious conclusion that we need to find stocks that move up in a nice and orderly fashion. (Location 1481)
We need to take both the momentum and the volatility into account. (Location 1483)
Try to start from scratch and design analytics for your own purpose, without using the usual technical analysis terminology. (Location 1487)
For measuring momentum, I use exponential regression. This opens the two obvious questions, what is regression and why is it exponential. (Location 1501)
Linear regression is a method of fitting a line over a series of values. It’s a way of finding the best fitting line, in this case to a time series of prices. (Location 1505)
The linear regression slope is therefore a measure of the speed, or momentum of the stock. (Location 1517)
If you annualize the slope, you get a number that tells you how much, in theory, the stock would gain in a full year if it continued the exact same angle. (Location 1526)
The exponential slope in this hypothetical case ended up at 0.0006. That means that on an average day, the stock moves up by 0.06% per day. (Location 1537)
Simple financial math tells us that 0.06% gains compounded for 250 days will end up at about 16% in a year. Now the number makes more intuitive sense. (Location 1540)
As mentioned, a neat part with this method of using annualized exponential regression slope is that it makes intuitive sense. We can see how many percent per year the current slope represents. (Location 1544)
In this book, we’re looking for a medium term momentum ranking. The regression calculations are all done using the past 90 trading days. (Location 1548)
There is still one little problem with our ranking approach. Using just annualized exponential regression for ranking means that we don’t care about the fit. (Location 1561)
The operative word here being ‘fit’. Since we’re using regression math, there’s a perfectly fine method to measure how well our price data fits the regression line. (Location 1570)
The R2 tells you how well the price series fits the regression line. If you have a bunch of random price points all over the place, you can still calculate a regression line. (Location 1573)
Zero is the minimum value for the R2 while one is the maximum. A value of one means that the data is a perfect fit to the line, and the lower the R2, the worse the fit of the regression line. (Location 1580)
What this means is that we measure the pure momentum, in the regression slope, and then we punish it for volatility. (Location 1588)
The difference is that the stocks with the most extreme fits, good or bad, will make large shifts in rank. (Location 1590)
Start buying from the top of the list, until you run out of cash. That’s how you make your initial portfolio. (Location 1738)
All in all, it’s the kind of portfolio that a skilled fundamental analyst might have constructed. It’s a portfolio I would be perfectly comfortable with and not lose any sleep over. (Location 1751)
This rule just makes sure that you don’t buy stocks that are moving sideways or down, just because there are no available stocks moving up. (Location 1758)
When it comes to position size, you need to remember that we’re not allocating money. We’re allocating risk. (Location 1776)
This is a very common mistake made not only by the vast majority of retail traders and investors but also by a large amount of fund and asset managers. (Location 1780)
Rebalancing is about how you change your position size over time. And no, I don’t mean how you double the size on success or how you double down when you lose. (Location 1949)
For longer term investment strategies such as equity momentum portfolios, a bi-weekly or monthly frequency is good enough. (Location 1985)
Well, sorry to disappoint you. There are no stop losses here. (Location 1997)
First check the market regime filter. Is the index trading above its 200 day moving average? If it’s not, you’re not allowed to buy. Yes, this means that you’ve got an automatic exposure scale-out. (Location 2039)
What we’re dealing with here is a long term method of beating the stock market. (Location 2073)
Because Wednesdays happen to have a 20% probability of being the best possible weekday to trade. (Location 2076)
Perhaps the whole momentum strategy seems a bit theoretical at this point. It might help if we take a look at the practical side and see how you would actually implement it. (Location 2124)
Obviously the first task here is to make sure that the S&P 500 is trading above the moving average. This part is easy. Any charting software can do that for you, and you could even do it in Excel if you want by just comparing the average of the last 200 index prices to the current one. (Location 2130)
The slope column is probably the most important one. That’s what the table is sorted on and that’s what governs the priority of our purchases. (Location 2139)
This is not a strategy where you can expect a steady 10% a year. Very few strategies are. It’s not a strategy where you can expect every year to be positive. (Location 2398)
We are after all buying stocks. Strategies based on buying stocks tend to look very similar in the long run. (Location 2400)
As Table 12‑1 shows, the momentum strategy yielded over 12% annualized over this 16 year period. You don’t think 12% per year is good enough? Think again. (Location 2427)
Most importantly though, we achieved our 12% at less than half the drawdowns of the index. Drawdown refers to the maximum loss we’ve seen during the time period. (Location 2435)
When writing my first book, Following the Trend, which is about trend following futures strategies, I struggled with how to best convey what it’s really like to be a professional asset manager. (Location 2567)
My solution in that book was to write a monster chapter where I explained, year by year, how that trading strategy had performed in the past. (Location 2572)
We start our portfolio just after the 1998 New Year’s party. We run the calculations, crunch the numbers, rank the stocks and build our portfolio. (Location 2603)
You can see in Figure 13‑2 how the cash started building up in October, as the index dipped under the water line. (Location 2690)
That’s usually quite a scary point. We had just closed out most positions, the markets are looking increasingly tired, there’s a potential bear market looming, and we’re now buying a full portfolio of stocks. (Location 2702)
When we enter into 2000 there are no real signs of a market reversal. We had a very long lasting bull market behind us and there was no reason to think that it might end soon. (Location 2734)
After that, most of the year went sideways. We started building up a portfolio again during the summer as the index came up a little bit and there were more available stocks to buy. This time however, the sector allocation looked very different. (Location 2836)
Discretionary, staples, energy, industrials, financials and materials were all represented. This is also the reason for the outperformance that we showed late this year. (Location 2841)
What we’ve seen so far is a fully working momentum investment model. The results over time are strong and it has stood the test of time. Bull or bear market, this model has shown great outperformance over time. (Location 4295)
thing and shouldn’t be confused with optimizations. (Location 4410)
Tags: orange