Mark Minervini (Meltdown)

[longandshort]

I'll try anything once.
What's the process? Remember to keep it simple.

Here i built this quick google sheet you can reference:


you can replace tickers and positions with your own and the data should calculate automatically.

all data using 35-day historicals

1 - gives you portfolio overview so you can see the risk your taking from a beta standpoint, expected returns, and expected daily standard deviation
2- factor exposures, you can set a target beta, and by changing your positions (#6) you'll get closer to your target
3- standard deviation so you can see what your daily fluctuations in price look like (2std moves occur about 95% of the time, though in stock prices 1std moves are approx 80%)
4- how long your trade is expected to last in days so you can eventually measure profit per day metrics (5% over 1 day might be better than 20% over 10 days)
5- your target price for the security
6- your position size
7- factor betas -- e.g. how sensitive each position is to a move in one of the factors
8- comparative data (vs. factors)

super basic but you can play with this to see how changes in your positions would impact portfolio level items. this is the basics of risk management.

i have not added a covariance analysis, but you can get a proxy of that by considering secrity correlations in the analysis tab. avoid having multiple high correlated positions.

if anyone wants access to the sheet ping me your email and i'll share it

 

[deaddog]

Here i built this quick google sheet you can reference:

Thanks. I'm on the road today but will get back to you.

 

[deaddog]

Here i built this quick google sheet you can reference:

I got the spreadsheet you emailed thank you for spending the time and effort.

1 - gives you portfolio overview so you can see the risk your taking from a beta standpoint, expected returns, and expected daily standard deviation

From what I gather I use this to assess my holdings. I have to input some assumptions such as target price, hold time and position size.

2- factor exposures, you can set a target beta, and by changing your positions (#6) you'll get closer to your target

I'm still having trouble figuring out how I measure factor exposure or even what it is. I understand that several factors contribute to price movement but how do I know or even guess which factor will move the price tomorrow?

3- standard deviation so you can see what your daily fluctuations in price look like (2std moves occur about 95% of the time, though in stock prices 1std moves are approx 80%)

I can see where this might be useful in determining position size.

4- how long your trade is expected to last in days so you can eventually measure profit per day metrics (5% over 1 day might be better than 20% over 10 days)

I don't usually have a set time. Hold time depends on how the price acts.

5- your target price for the security
6- your position size

These are assumptions that I calculate and input.

7- factor betas -- e.g. how sensitive each position is to a move in one of the factors
8- comparative data (vs. factors)

See #2. above

super basic but you can play with this to see how changes in your positions would impact portfolio level items. this is the basics of risk management.

I think spreadsheets are great. A what if tool. What if I change this , what happens to my total return. I do play with them from time to time. But they are based on assumptions, the results will only be accurate if your assumptions are correct.

Until I figure out what a factor beta is (and I promise I'll keep trying) There isn't much sense in me making assumptions on what they might be.

As an aside my first experience with spreadsheets and investing was a financial advisor (mutual fund salesman) showing me how my investment would grow at different percent returns over the years. Never once mentioned negative years.

 

[longandshort]

From what I gather I use this to assess my holdings. I have to input some assumptions such as target price, hold time and position size.

Correct

I'm still having trouble figuring out how I measure factor exposure or even what it is. I understand that several factors contribute to price movement but how do I know or even guess which factor will move the price tomorrow?

This sheet measures factor exposure for you using a historical 35-day analysis.

I don't usually have a set time. Hold time depends on how the price acts.

If on average you hold a stock for X days, use that. If you're not sure, use 20 (approx 1 month)

I think spreadsheets are great. A what if tool. What if I change this , what happens to my total return. I do play with them from time to time. But they are based on assumptions, the results will only be accurate if your assumptions are correct.

True but this sheet tells gives you your factor exposure. You're not making any assumptions in this spreadsheet. You can make assumptions outside of it -- such as a desire to hold onto momentum, and use this spreadsheet to see how adding another position would change your momentum exposure.

Until I figure out what a factor beta is (and I promise I'll keep trying) There isn't much sense in me making assumptions on what they might be.

it's the systematic return of things like momentum and value, etc.

As an aside my first experience with spreadsheets and investing was a financial advisor (mutual fund salesman) showing me how my investment would grow at different percent returns over the years. Never once mentioned negative years.

Well this sheet isn't designed to tell you how much you can make. It's meant to show you what your positions look like from an expected return, factor exposure, and portfolio standpoint so you can make useful decisions about risk.

 

[deaddog]

Well this sheet isn't designed to tell you how much you can make. It's meant to show you what your positions look like from an expected return, factor exposure, and portfolio standpoint so you can make useful decisions about risk.

I gotta be wearing your patience thin and for that I'm appreciative.

So I input the stock I have picked along with my position size and a wild-assed guess for a target. How does that tell me my factor exposure?

The stock was chosen because it has momentum. (All my stocks are chosen that way) My risk is more or less calculated based on a percentage of my account. Isn't the majority of my factor exposure momentum?

Sure there is market risk, if the market dives usually all stock follow but that's where my stops come into play. That's what kept me out of the 2008 and 2020 downturns.

I don't see any other way to adjust my risk. I suppose I could play with my stop or my target or position size to try and come up with some optimal risk but those models never seem to play out in real time.

Does knowing my expected return and factor exposure have any affect on what the price will do next week?

 

[longandshort]

I gotta be wearing your patience thin and for that I'm appreciative.

So I input the stock I have picked along with my position size and a wild-assed guess for a target. How does that tell me my factor exposure?

You don't need a price target for factor exposure. To assess factor exposure, add a ticker. The sheet will automatically pull up historical stock data and compare it against the benchmark for the factor to calculate factor beta (proxy for exposure here).

The stock was chosen because it has momentum. (All my stocks are chosen that way) My risk is more or less calculated based on a percentage of my account. Isn't the majority of my factor exposure momentum?

Your goal is to try to add momentum stocks to your portfolio. This sheet tells you how each stock contributes to that factor (and others). Generally, since you are "factor loading", you want your momentum factor exposure to be high, but limit exposure to other factors.

Sure there is market risk, if the market dives usually all stock follow but that's where my stops come into play. That's what kept me out of the 2008 and 2020 downturns.

Ya this is not a sheet for stop losses lol. This sheet tells you your factor exposure and risk contributions -- so you can identify where your risk is actually coming from.

I don't see any other way to adjust my risk. I suppose I could play with my stop or my target or position size to try and come up with some optimal risk but those models never seem to play out in real time.

yeah this is not for stop losses

Does knowing my expected return and factor exposure have any affect on what the price will do next week?

No.

Use this to figure out what would happen if momentum went down 1% -- your portfolio beta to momentum would translate into a return figure.

E.g. if your portfolio beta to momentum is 2, then for every 1% decline in momentum, you would experience 2%.

 

[deaddog]

You don't need a price target for factor exposure. To assess factor exposure, add a ticker. The sheet will automatically pull up historical stock data and compare it against the benchmark for the factor to calculate factor beta (proxy for exposure here).


Your goal is to try to add momentum stocks to your portfolio. This sheet tells you how each stock contributes to that factor (and others). Generally, since you are "factor loading", you want your momentum factor exposure to be high, but limit exposure to other factors.

My stock selection process by default should select stocks with the highest momentum factor.
Stocks making new highs. Stacked moving averages. Ranked by outperformance of the benchmark index.

Use this to figure out what would happen if momentum went down 1% -- your portfolio beta to momentum would translate into a return figure.

E.g. if your portfolio beta to momentum is 2, then for every 1% decline in momentum, you would experience 2%.

OK so what do I do if momentum went down 1%. It is interesting information but how does it help me manage my portfolio.

My process right now is to scan for stocks that have high but steady momentum.
Stocks making new highs
Rank by outperformance of benchmark
Determine position size based on percentage of account and stop loss placement.
Start at the top of the list and buy until I run out of cash.
Wait until a stock hits an exit.
Repeat ranking process.
Buy the stocks I don't own until I run out of cash.

I throw in a few other rules as to when I can buy. I don't buy anything if the benchmark index is trading below 150 day ma. I don't buy on days the market is trending downwards.

Given that process how is knowing what will happen if the momentum went down 2% going to help?

 

[David's faith]

Here i built this quick google sheet you can reference:


you can replace tickers and positions with your own and the data should calculate automatically.

View attachment 269952

all data using 35-day historicals

1 - gives you portfolio overview so you can see the risk your taking from a beta standpoint, expected returns, and expected daily standard deviation
2- factor exposures, you can set a target beta, and by changing your positions (#6) you'll get closer to your target
3- standard deviation so you can see what your daily fluctuations in price look like (2std moves occur about 95% of the time, though in stock prices 1std moves are approx 80%)
4- how long your trade is expected to last in days so you can eventually measure profit per day metrics (5% over 1 day might be better than 20% over 10 days)
5- your target price for the security
6- your position size
7- factor betas -- e.g. how sensitive each position is to a move in one of the factors
8- comparative data (vs. factors)

super basic but you can play with this to see how changes in your positions would impact portfolio level items. this is the basics of risk management.

i have not added a covariance analysis, but you can get a proxy of that by considering secrity correlations in the analysis tab. avoid having multiple high correlated positions.

if anyone wants access to the sheet ping me your email and i'll share it

pinged you my mail in a private thread. Thanks for sharing

 

[longandshort]

My stock selection process by default should select stocks with the highest momentum factor.
Stocks making new highs. Stacked moving averages. Ranked by outperformance of the benchmark index.

I get that -- so technically, you should expect that your portfolio is very correlated to the momentum benchmark. Correlations are time varying, so you should expect the correlation to change over time. It is useful for you to know how correlated you are to momentum because you can establish factor drift limits. E.g. if your portfolio correlation to momentum, drill down to see what position is causing this and whether you are comfortable with the risk or not.

OK so what do I do if momentum went down 1%. It is interesting information but how does it help me manage my portfolio.

Portfolio theory tells us that you want to allocate your portfolio in such a way that it either generates the highest risk/reward (sharpe ratio) or lowest variance. For any random 5 stocks there is an "optimal" portfolio weighting, using expected returns, standard deviation, and a covariance-variance matrix. I did not include the last piece in the spreadsheet.

My process right now is to scan for stocks that have high but steady momentum.
Stocks making new highs
Rank by outperformance of benchmark
Determine position size based on percentage of account and stop loss placement.
Start at the top of the list and buy until I run out of cash.
Wait until a stock hits an exit.
Repeat ranking process.
Buy the stocks I don't own until I run out of cash.

this spreadsheet will help you figure out what mix of positions will lead to 1) greatest factor loading, 2) optimal portfolio weights for max expected sharpe (not solved, manual entry, I don't know if google has a solver lol)

The idea is that as a portfolio manager of your own account, you should be overweight where you think you can make more money, and underweight where you are less certain (high standard deviation, or lower expected return) of making money.

I added a factor model to generate expected returns. It's not super robust but you can compare it ex-post to see why returns were different and what caused them to deviate.

Given that process how is knowing what will happen if the momentum went down 2% going to help?

Because some of your positions are contributing to volatility and momentum than others. Typically a naïve portfolio will be equal-weight, all other portfolios are expressing a relative view on one stock vs the others.

 

[deaddog]

Portfolio theory tells us that you want to allocate your portfolio in such a way that it either generates the highest risk/reward (sharpe ratio) or lowest variance. For any random 5 stocks there is an "optimal" portfolio weighting, using expected returns, standard deviation, and a covariance-variance matrix. I did not include the last piece in the spreadsheet.


this spreadsheet will help you figure out what mix of positions will lead to 1) greatest factor loading, 2) optimal portfolio weights for max expected sharpe (not solved, manual entry, I don't know if google has a solver lol)

How do you recommend I change my process to accomplish this.