How do you use Portfolio123 (or anything else) to track which stocks are the "best" within each factor, so you know which ones to track for slowing momentum in that factor?
my ranking system has 7 factor groups and they are weighted. So if the composite out of that weights fall below the sell threashold, the system fires a sell signal, so its kind of a rank factor stop loss… Also interesting is, that if you have a buy rule for example Ramk > 99 the system might not find a new stock int thr illiqiud small cap space and them it raises cash because maybe at the same time some stocks fall under the sell threashold. Build in market timing!
Thanks! And within those factors, eg High Beta or China ADRs or even within Rank Momentum, how do you set it up to track the top stocks within each factor? Because the ETFs are too muddled to cleanly present the factor to you
Well, the first step is to find out what factors are always a good choice. High Beta not, because High beta has good trends, but if it tanks it can take a decade before it comes back. It can loose up to 95% like in 2000 and it took to 2013 to get all back).
So, for example value + momentum (especially in illiquid small caps) is doing well in 3.5 Regimes (Goldi, Refla., Stagflation (esp. with higher inflation) and a shallow deflation). It only tanks (up to 50%) when we get a deep deflation. But it comes back extremely fast, after a 50% DD it takes almost never more than a year until that DD is gone, and the system makes a new high. Most of my systems where back to ATHs mid-2009, that is a really, really good behavior!
So, one approach could be, that you simply go long factors that are almost doing fine (which Buffett is doing: big caps Quality, Value, reasonable growth, good management quality (yep is a factor!), low volatility is doing pretty well (on average) in all regimes, with again a hefty drawdown every 10 Years, remember he cannot go long small caps!). Buffett is doing the 100% right thing for a very big portfolio!
Now the second steps comes. Which factors do trend well, e.g., for example high beta! When high beta is going up with relative strength to the market for 4 Weeks (also works 3-9 months), you can go long that factor (or short if weak, like now!). There are only 5-7 Factors that trend well: High Beta one of the best (that is the reason why CTA Trend Follower trade that exposure, it trends as well as many commodities), Small Size, Industry Momentum, 12 Month momentum for example as well, Country Momentum. Value, not really good, but for example international Value is trending really well (See https://www.aqr.com/Insights/Research/Working-Paper/Factor-Momentum-Everywhere). So, you can trend follow those factors. Other factors like Quality do not trend well but you can be long them all the time.
Third step is to find out which of the trending factors does give hints for the regime. If growthy high beta is doing fine you most likely are in a reflation or goldilocks regime, if not you are most likely in a stagflation or deflation regime (makes sense if you look at how interest rates behave in stagflation (mostly up) and deflation (first mostly up which leads to deflation).
For this I build clean factor systems: High Beta, China ADR, tons of stuff around small cap value momentum, Quality stuff etc.
China ADRs: they trend hard and fast (up and down) and academics found out about Country momentum factor. Country momentum factors work, but they work best if the Country consist of people who are not very individualistic (like China!). Also, from looking at the capital curves of my China systems that can be very much seen. From what I saw in the data, China ADRs are really trashy (bad earnings, high volatility, high beta etc.). But when they go and go together with High Beta you can be long as hell (does not take very long though, like from November 2020 Feb. 2021), this is the time where risk takers make 100s of % (if they know how to time this, otherwise they lose everything and more on the trend reversal). China ADRs are the first to drop after a burst up with high beta no earnings (around March 2021).
So, what I basically have done is to build systems that have a clean factor loading.
I then put them into a regime grid to know where we are.
I then be long the factors that have monster great advantages for small portfolios (sup 1 Mill lets say) which is (you guessed it) illiquid small cap value momentum (5 System with some variation on Quality, EPS Revisions, industry momentum. Those factors almost always work well so I am diversified). In order to mitigate risk in a stagflation regime (where illiquid small cap momentum stocks do fine, but are very volatile and deflation is always a risk if you are in stagflation) or deflation regime I use my Grid regime (+Data from 42Macro as a second opinion) to find out when my long small cap value momentum stuff needs to be hedged (like right now).
What find out is: 1. What factors give a small account a big advantage. 2. What factors work well almost all the time. 3. How can big drawdowns be mitigated (e.g. when I have to be long / Short). For this the Regime GRID. Actually step 3. Would not be necessary, because small cap value momentum comes back extremely fast, but if you got a 50% DD, you simply cannot think straight anymore. Therefore the 3. Step.
Sorry for the long answer, I will write a blog post about it here as well.
1. Do you ever get concerned that we are training these factors on datasets that only include falling interest rates (1982-on), which may not be the case going forward?
2. When I think about when smallcaps underperforming, I think about illiquidity runs -- taking out factor-speak, a lot of these smallcaps are trading as call options on their credit structures. Does a lot of your implementation come down to *timing* when the Fed has gone too far and tipped us into a less liquid regime?
3. When you build a factor following the top performers in some area, eg Industry Momentum or China ADRs, do you have video/post showing exactly how we could do the same thing on portfolio123? Would love to play around with similar ideas (and would totally understand if you did so for premium subs if you were up for it)
1. --> My backtests go for 20 years. But I have read tons of academic papers which go back to the 1870s (usually they go back to the 1960s).
The results are the same. Also this environmet whe are in matches 2000-2003 very well.
2. Yes, small cap value momentum will tank in a deflation regime, which will ultimatly come. For this we are long short. Right now we win on both sides (long and short). In Deflation we will be flat with the port (performanc wise). But on the bottom small caps will jump 10% in one day and it takes a week or so to get in. The hedge can be reduced in seconds. Also in the secon half of deflation small cap value momentum is doing very well!
3. Yes, will do this in a video (might take some time, but I will deliver!!!)
It seems like the market is pricing in more deflation hence risk assets are heading down together. I'm personally waiting to put back my shorts by the end of the week or after Friday's OPEX. I'm leaning toward shorting QQQ instead of IWC since the latter is picking up relative strength (RS) against the QQQ. Do you use any relative strength metric or sticking to profolio.com system? Thank you!
I like IWM because it correlates very well with my small cap longs. QQQ fine as well, if we tank it will not make a big difference, QQQ has higher beta, e.g. the swings are bigger...
How do you use Portfolio123 (or anything else) to track which stocks are the "best" within each factor, so you know which ones to track for slowing momentum in that factor?
my ranking system has 7 factor groups and they are weighted. So if the composite out of that weights fall below the sell threashold, the system fires a sell signal, so its kind of a rank factor stop loss… Also interesting is, that if you have a buy rule for example Ramk > 99 the system might not find a new stock int thr illiqiud small cap space and them it raises cash because maybe at the same time some stocks fall under the sell threashold. Build in market timing!
Thanks! And within those factors, eg High Beta or China ADRs or even within Rank Momentum, how do you set it up to track the top stocks within each factor? Because the ETFs are too muddled to cleanly present the factor to you
Well, the first step is to find out what factors are always a good choice. High Beta not, because High beta has good trends, but if it tanks it can take a decade before it comes back. It can loose up to 95% like in 2000 and it took to 2013 to get all back).
So, for example value + momentum (especially in illiquid small caps) is doing well in 3.5 Regimes (Goldi, Refla., Stagflation (esp. with higher inflation) and a shallow deflation). It only tanks (up to 50%) when we get a deep deflation. But it comes back extremely fast, after a 50% DD it takes almost never more than a year until that DD is gone, and the system makes a new high. Most of my systems where back to ATHs mid-2009, that is a really, really good behavior!
So, one approach could be, that you simply go long factors that are almost doing fine (which Buffett is doing: big caps Quality, Value, reasonable growth, good management quality (yep is a factor!), low volatility is doing pretty well (on average) in all regimes, with again a hefty drawdown every 10 Years, remember he cannot go long small caps!). Buffett is doing the 100% right thing for a very big portfolio!
Now the second steps comes. Which factors do trend well, e.g., for example high beta! When high beta is going up with relative strength to the market for 4 Weeks (also works 3-9 months), you can go long that factor (or short if weak, like now!). There are only 5-7 Factors that trend well: High Beta one of the best (that is the reason why CTA Trend Follower trade that exposure, it trends as well as many commodities), Small Size, Industry Momentum, 12 Month momentum for example as well, Country Momentum. Value, not really good, but for example international Value is trending really well (See https://www.aqr.com/Insights/Research/Working-Paper/Factor-Momentum-Everywhere). So, you can trend follow those factors. Other factors like Quality do not trend well but you can be long them all the time.
Third step is to find out which of the trending factors does give hints for the regime. If growthy high beta is doing fine you most likely are in a reflation or goldilocks regime, if not you are most likely in a stagflation or deflation regime (makes sense if you look at how interest rates behave in stagflation (mostly up) and deflation (first mostly up which leads to deflation).
For this I build clean factor systems: High Beta, China ADR, tons of stuff around small cap value momentum, Quality stuff etc.
China ADRs: they trend hard and fast (up and down) and academics found out about Country momentum factor. Country momentum factors work, but they work best if the Country consist of people who are not very individualistic (like China!). Also, from looking at the capital curves of my China systems that can be very much seen. From what I saw in the data, China ADRs are really trashy (bad earnings, high volatility, high beta etc.). But when they go and go together with High Beta you can be long as hell (does not take very long though, like from November 2020 Feb. 2021), this is the time where risk takers make 100s of % (if they know how to time this, otherwise they lose everything and more on the trend reversal). China ADRs are the first to drop after a burst up with high beta no earnings (around March 2021).
So, what I basically have done is to build systems that have a clean factor loading.
I then put them into a regime grid to know where we are.
I then be long the factors that have monster great advantages for small portfolios (sup 1 Mill lets say) which is (you guessed it) illiquid small cap value momentum (5 System with some variation on Quality, EPS Revisions, industry momentum. Those factors almost always work well so I am diversified). In order to mitigate risk in a stagflation regime (where illiquid small cap momentum stocks do fine, but are very volatile and deflation is always a risk if you are in stagflation) or deflation regime I use my Grid regime (+Data from 42Macro as a second opinion) to find out when my long small cap value momentum stuff needs to be hedged (like right now).
What find out is: 1. What factors give a small account a big advantage. 2. What factors work well almost all the time. 3. How can big drawdowns be mitigated (e.g. when I have to be long / Short). For this the Regime GRID. Actually step 3. Would not be necessary, because small cap value momentum comes back extremely fast, but if you got a 50% DD, you simply cannot think straight anymore. Therefore the 3. Step.
Sorry for the long answer, I will write a blog post about it here as well.
Thanks! 3 quick follow-ups:
1. Do you ever get concerned that we are training these factors on datasets that only include falling interest rates (1982-on), which may not be the case going forward?
2. When I think about when smallcaps underperforming, I think about illiquidity runs -- taking out factor-speak, a lot of these smallcaps are trading as call options on their credit structures. Does a lot of your implementation come down to *timing* when the Fed has gone too far and tipped us into a less liquid regime?
3. When you build a factor following the top performers in some area, eg Industry Momentum or China ADRs, do you have video/post showing exactly how we could do the same thing on portfolio123? Would love to play around with similar ideas (and would totally understand if you did so for premium subs if you were up for it)
Very good questions!!!
1. --> My backtests go for 20 years. But I have read tons of academic papers which go back to the 1870s (usually they go back to the 1960s).
The results are the same. Also this environmet whe are in matches 2000-2003 very well.
2. Yes, small cap value momentum will tank in a deflation regime, which will ultimatly come. For this we are long short. Right now we win on both sides (long and short). In Deflation we will be flat with the port (performanc wise). But on the bottom small caps will jump 10% in one day and it takes a week or so to get in. The hedge can be reduced in seconds. Also in the secon half of deflation small cap value momentum is doing very well!
3. Yes, will do this in a video (might take some time, but I will deliver!!!)
Thank you!!!! :-)
It seems like the market is pricing in more deflation hence risk assets are heading down together. I'm personally waiting to put back my shorts by the end of the week or after Friday's OPEX. I'm leaning toward shorting QQQ instead of IWC since the latter is picking up relative strength (RS) against the QQQ. Do you use any relative strength metric or sticking to profolio.com system? Thank you!
RS between QQQ and IWM adusted with beta is pretty bad in both..., hard to decide...
I like IWM because it correlates very well with my small cap longs. QQQ fine as well, if we tank it will not make a big difference, QQQ has higher beta, e.g. the swings are bigger...