I'm curious what do you use to create a momentum investing portfolio? my hope is we get many responses so we can share,compare and hopefully improve each others momentum plans.
Hello I am 30 year old tram driver from Manchester.Manchester Momentum Investors facebook group
'Proceeds of any sales should be reinvested in the best IT not in the portfolio that satisfies the criteria'
once you sell an Investment trust do you buy a new one to add to the portfolio or buy more of the best performing investment trust in your portfolio?
Once I sell an IT, I buy a new one that's not in the portfolio so the total number in the portfolio stays at ten.
Further to my previous post, I've tried to keep my momentum strategy with ITs as easy to execute as possible without too many trades, so each time I sell an IT I reinvest the whole proceeds into the new IT that I buy. This means my fantasy IT portfolio has moved a long way from equal weighting of its ten constituents over the 14 months it has been in existence. However, if the portfolio eventually gets so unbalanced that I sell an IT and its proceeds represent more than 15% of the portfolio's value, my plan is to use that cash to invest 10% of the portfolio's value in the new IT and use whatever cash is left over to invest in the IT with the smallest weighting in the portfolio to bring its weighting closer to 10%. So I do plan to do some rebalancing eventually, but only occasionally and in an incomplete way.
I do something like this with my real IT portfolio, in which (as I wrote in a previous post) I take both share price momentum and discounts to NAV into account and I also try to keep a reasonable level of diversification across different types of IT, so I use some (imperfect) judgement as well as rules.
In my real IT portfolio, I also aim to have about 30% of its total value in ITs with an 'absolute return' approach, in the hope that this will reduce the drawdown of the portfolio's total value in a bear market although it reduces returns in a bull market, but I only rebalance back to that 30% target on an occasional basis (for example when I add new cash to the portfolio). The 'absolute return' ITs I use for this purpose at present are Capital Gearing, Personal Assets and the Ruffer Investment Company.
Hi Chrisopher,
I've just watched the "Magic of Momentum Trading" video you put a link to above. It has some interesting ideas, so thanks for the link.
Hi David,
You wrote:
"I decided the easiest way of creating a momentum strategy was just to pick shares with the highest momentum ranks on the basis that Stockopedia had done all the hard work on pulling together research done by the dozens of market professionals and academics who have really studied this. I then exclude larger companies, on the basis that factor investing works best for smaller companies, exclude those that I consider untradable due to large spreads, and put a minimum quality filter or value filter on, to avoid 'momentum traps'. I prefer steady consistent rises in share price, not large occasional jumps too."
This seems to me to be a good approach. I would be interested to know whether you sell all the stocks in your momentum portfolio on a periodic basis (e.g. every 3, 6 or 12 months) and reinvest the proceeds in a new portfolio, or whether you sell a stock if its Momentum Rank falls below some pe-determined threshold (e.g. 70) and reinvest the proceeds in a new stock.
I buy top stocks each month, then sell after 12 months unless the business changes. I did panic sell two stocks after the share price dropped heavily, and if I had held, would actually have been among my top performers, so I am trying to hold my nerve now. This has a lot of similarities with NAPS approach, but done monthly, with momentum being the main driver.
I don't think I am smart enough, or emotionally stable enough to a stockpicker, tried that and it did not go well!
NAPS 2021 is up 32.6% this year as a comparison, would be interested to see if your Momentum stocks are better or worse.
Yes, it would, I am still working full time, so not able to do all the analysis I would like to, but that would an interesting comparison.
If anybody is interested in bringing a ZOOM group together for momentum through SIGnet, then that is something we could organise.
great, have you heard of the salty dog investor? I don't use it but I believe it's about picking funds with high momentum also
Hi Chris,
I have heard of the Salty Dog Investor web site, which charges users to recommend funds with momentum. FundExpert is a similar web site, which also charges users but has different criteria for recommending funds with momentum. I don't use either web site, because both restrict themselves to recommending OEIC funds, and also because there are a number of web sites that provide free past performance data for OEICs.
Personally, I don't think OEICs are ideal for a momentum strategy using funds, because it can take a couple of days to sell an OEIC and a further day or two to reinvest the proceeds in another OEIC, and I don't like waiting in that way. Investment trusts and ETFs have the advantage over OEICs that you can sell one and reinvest the proceeds in another within minutes. ITs have the disadvantage that most incur 0.5% stamp duty when they are bought, which ETFs and most OEICs don't. Some ITs also have the disadvantage that they have wide bid-offer spreads, which is why I try to restrict my IT purchases to those with low spreads. But ITs have the advantage over ETFs that they can be at significant discounts to Net Asset Value, so if one buys such ITs which are also showing recent share price momentum then there is a reasonable chance that their discounts will narrow which would improve their future share price performances. Of course, in a bear market IT's discounts tend to widen, so ITs can be more volatile and riskier than OEICs and ETFs.
I have now watched again the "Magic of Momentum Trading" video that Christopher put a link to above. Here is another link to it:
https://www.youtube.com/watch?v=TjZ8JlNk0xU
The video is a one-hour interview of a momentum trader called Alan Clement by Andrew Swanscott of the Better System Trader website. I think Alan Clement presents some good ideas for systematic momentum trading, and I wondered how well one could replicate those ideas using the information provided by Stockopedia. Here is my summary of what I think are the main points he makes and how one might implement them using Stockopedia:
- He says momentum trading is profitable on average because there is a tendency for share price trends to persist for some time. The stocks with the strongest trends are likely to be the most profitable, but one needs to buy a portfolio of stocks, (a) because there is no guarantee that any individual stock purchase will be profitable and there tends to be an 80:20 rule, i.e. 80% of profit comes from 20% of positions, and (b) to reduce the volatility of the portfolio.
- He thinks weekly monitoring of his portfolio is about the right frequency, for example every Monday.
- One needs some way to measure trends, in order to identify opportunities, and also to decide when to sell a stock from the portfolio. He mentions "percentage change in price" or "numbers of ATRs". I think by ATR he means Average True Range, which is a technical indicator not provided by Stockopedia. But Stockopedia does provide valus for "percentage change in price" over a number of different timescales.
- He likes to use smoothed price indicators. He says one way to do this would be to use a percentage exponential moving average. He doesn't like ordinary moving averages, because a 5 dollar rise in a 5 dollar share price is very different from a 5 dollar price on 100 dollar share price. But a percentage exponential or ordinary moving average would not suffer from that objection. Percentage exponential moving averages are not available on Stockopedia, but percentage moving averages are available over various timescales.
- He thinks a 6-month look-back period for his trend indicator is in what he calls the "sweet spot". Stockopedia provides a 130-day percentage moving average (%130dMA), which it describes as being roughly equivalent to a 6-month percentage moving average because the 130 days refers to the number of trading days being considered. A positive value of %130dMA means the stock is trending up over 6 months, a negative value means it is trending down over that period.
- He likes to have between 12 and 15 stocks in his momentum portfolio, with roughly equal initial weighting in each stock, but he likes to allow any winners to run without rebalancing them to equal weighting. One approximate way to achieve this with, say, 15 stocks, would be to start off with an initial equal weighting of 6.7% in each stock. When a stock is sold, if the proceeds are less than 6.7% of the portfolio's total value, then all that cash would be invested in whatever new stock is bought. But if the proceeds are more than 6.7%, then 6.7% of the total value would be invested in the new stock and the remaining cash would be invested in the existing position that has the lowest value to bring it up to 6.7%, and so on if necessary.
- He does not think weighting by a stock's volatility is helpful to the portfolio's performance.
- He sells a stock at the time of each weekly review if its trend indicator is significantly worse than that of the best stock in the candidate universe that is not currently in the portfolio. For example, if one is using the %200dMA as the indicator, then one might sell a stock if its %200dMA is, say, 20% below that of the best candidate for entry. [I don't think he discusses this in the interview, but I think one needs to have some minimum threshold, such as 20%, to prevent over-trading and excessive execution expenses.]
- He likes to use the same indicator for both entry and exit of a position, and he likes the concept of Dual Momentum, which is to use both "relative momentum" of the stocks within a particular "universe" relative to each other, and the "absolute momentum" of the stock price relative to its own previous history. He does not invest in a stock or will sell it from his portfolio if its price is below a particular "absolute momentum" threshold. A simple "absolute momentum" rule would be to not invest in or to sell a stock if its %200dMA is negative.
- He does not use stops to decide when to sell stocks. Instead, he uses his chosen "absolute momentum" rule. For example, one might sell any stock if its %200dMA is negative. In that case, one would reinvest the proceeds in the stock not currently in the portfolio with the highest value of %200dMA. If there is no such stock in the universe being considered, then the proceeds would be kept as cash. This means the portfolio would gradually exit into cash during a bear market as more and more stocks in its universe have negative values of %200dMA. He thinks this is a very good way to control risk in his strategy and he calls this the "magic of momentum trading". When market conditions improve, his "absolute momentum" rules would mean that the portfolio would gradually increase the number of stock positions it holds until it reaches the desired number, e.g. 15, again.
I like Alan Clement's ideas, and indeed there are some similarities with the rules I described in a previous post for a fantasy momentum portfolio I created on Stockopedia using investment trusts.
I think the simplest way to implement these ideas in Stockopedia would be to create a screen with the following rules:
%130dMA > 0, sort by descending %130dMA
and to use that screen to select an initial equally weighted portfolio of however may stocks one would like to hold in it, and to monitor that portfolio once a week using the same screen as described above.
One could combine the above rules in one's screen with whatever other criteria one likes to use, if one thinks that would improve the profitability of the strategy. For example, if one wants to keep trading costs low, one could have a rule that bid-offer spreads must be less than a threshold such as 100 basis points (i.e. 1% of the share price). One could also use Stockopedia's StockRank as a criterion, for example StockRank > 80. Or one could restrict oneself to Stockopedia's "Super Stocks". One could also enforce some diversification in the portfolio, for example by not allowing more than one stock from any particular Stockopedia "Industry Group" when selecting new stocks for the portfolio.
I would be interested in other people's views on these ideas.
I am reposting, as a new comment, what I wrote in reply to an earlier comment, because I have just noticed that on the new site my new post does not appear as the latest comment so people might not notice it. (Until today, I have only used the old site which behaves differently from the new one.)
I have now watched again the "Magic of Momentum Trading" video that Christopher put a link to above. Here is another link to it:
https://www.youtube.com/watch?v=TjZ8JlNk0xU
The video is a one-hour interview of a momentum trader called Alan Clement by Andrew Swanscott of the Better System Trader website. I think Alan Clement presents some good ideas for systematic momentum trading, and I wondered how well one could replicate those ideas using the information provided by Stockopedia. Here is my summary of what I think are the main points he makes and how one might implement them using Stockopedia:
- He says momentum trading is profitable on average because there is a tendency for share price trends to persist for some time. The stocks with the strongest trends are likely to be the most profitable, but one needs to buy a portfolio of stocks, (a) because there is no guarantee that any individual stock purchase will be profitable and there tends to be an 80:20 rule, i.e. 80% of profit comes from 20% of positions, and (b) to reduce the volatility of the portfolio.
- He thinks weekly monitoring of his portfolio is about the right frequency, for example every Monday.
- One needs some way to measure trends, in order to identify opportunities, and also to decide when to sell a stock from the portfolio. He mentions "percentage change in price" or "numbers of ATRs". I think by ATR he means Average True Range, which is a technical indicator not provided by Stockopedia. But Stockopedia does provide valus for "percentage change in price" over a number of different timescales.
- He likes to use smoothed price indicators. He says one way to do this would be to use a percentage exponential moving average. He doesn't like ordinary moving averages, because a 5 dollar rise in a 5 dollar share price is very different from a 5 dollar price on 100 dollar share price. But a percentage exponential or ordinary moving average would not suffer from that objection. Percentage exponential moving averages are not available on Stockopedia, but percentage moving averages are available over various timescales.
- He thinks a 6-month look-back period for his trend indicator is in what he calls the "sweet spot". Stockopedia provides a 130-day percentage moving average (%130dMA), which it describes as being roughly equivalent to a 6-month percentage moving average because the 130 days refers to the number of trading days being considered. A positive value of %130dMA means the stock is trending up over 6 months, a negative value means it is trending down over that period.
- He likes to have between 12 and 15 stocks in his momentum portfolio, with roughly equal initial weighting in each stock, but he likes to allow any winners to run without rebalancing them to equal weighting. One approximate way to achieve this with, say, 15 stocks, would be to start off with an initial equal weighting of 6.7% in each stock. When a stock is sold, if the proceeds are less than 6.7% of the portfolio's total value, then all that cash would be invested in whatever new stock is bought. But if the proceeds are more than 6.7%, then 6.7% of the total value would be invested in the new stock and the remaining cash would be invested in the existing position that has the lowest value to bring it up to 6.7%, and so on if necessary.
- He does not think weighting by a stock's volatility is helpful to the portfolio's performance.
- He sells a stock at the time of each weekly review if its trend indicator is significantly worse than that of the best stock in the candidate universe that is not currently in the portfolio. For example, if one is using the %200dMA as the indicator, then one might sell a stock if its %200dMA is, say, 20% below that of the best candidate for entry. [I don't think he discusses this in the interview, but I think one needs to have some minimum threshold, such as 20%, to prevent over-trading and excessive execution expenses.]
- He likes to use the same indicator for both entry and exit of a position, and he likes the concept of Dual Momentum, which is to use both "relative momentum" of the stocks within a particular "universe" relative to each other, and the "absolute momentum" of the stock price relative to its own previous history. He does not invest in a stock or will sell it from his portfolio if its price is below a particular "absolute momentum" threshold. A simple "absolute momentum" rule would be to not invest in or to sell a stock if its %200dMA is negative.
- He does not use stops to decide when to sell stocks. Instead, he uses his chosen "absolute momentum" rule. For example, one might sell any stock if its %200dMA is negative. In that case, one would reinvest the proceeds in the stock not currently in the portfolio with the highest value of %200dMA. If there is no such stock in the universe being considered, then the proceeds would be kept as cash. This means the portfolio would gradually exit into cash during a bear market as more and more stocks in its universe have negative values of %200dMA. He thinks this is a very good way to control risk in his strategy and he calls this the "magic of momentum trading". When market conditions improve, his "absolute momentum" rules would mean that the portfolio would gradually increase the number of stock positions it holds until it reaches the desired number, e.g. 15, again.
I like Alan Clement's ideas, and indeed there are some similarities with the rules I described in a previous post for a fantasy momentum portfolio I created on Stockopedia using investment trusts.
I think the simplest way to implement these ideas in Stockopedia would be to create a screen with the following rules:
%130dMA > 0, sort by descending %130dMA
and to use that screen to select an initial equally weighted portfolio of however may stocks one would like to hold in it, and to monitor that portfolio once a week using the same screen as described above.
One could combine the above rules in one's screen with whatever other criteria one likes to use, if one thinks that would improve the profitability of the strategy. For example, if one wants to keep trading costs low, one could have a rule that bid-offer spreads must be less than a threshold such as 100 basis points (i.e. 1% of the share price). One could also use Stockopedia's StockRank as a criterion, for example StockRank > 80. Or one could restrict oneself to Stockopedia's "Super Stocks". One could also enforce some diversification in the portfolio, for example by not allowing more than one stock from any particular Stockopedia "Industry Group" when selecting new stocks for the portfolio.
I would be interested in other people's views on these ideas.
Thanks for the details. With reference to ATR, this is provided in the Chart Section for each security, under Indicators.
Thanks for the info about ATR, Wilkonz.
I forgot to mention that Alan Clement decribes his momentum trading strategy as a "rotational" one.
A couple of advantages of using funds (i.e.OEICs, ITs or ETFs) over individual stocks in a momentum strategy are:
1. They are much less volatile than individual stocks, so one needs far fewer funds than stocks in a portfolio for a particular level of risk.
2. It is almost impossible to lose one's entire investment in a fund, but that is possible with an individual stock if it goes out of business or turns out to be a fraud (e.g. Globo a few years ago).
If one would like to use ETFs in a momentum strategy, unfortunately Stockopedia provides very little information about them. The best website I have found so far for information about ETFs is:
The JustETF website provides past performance data in terms of percentage return over various time periods, such as 3, 6 or 12 month, which could be used in a momentum strategy. However it doesn't provide percentage moving average data, which I prefer to use if available. But it does allow one to order ETFs by one-year return / risk, which might be a useful indicator for a momentum strategy.
