OK. But to do that do I not have to make certain assumptions and the results the spreadsheet calculates are based on those assumptions. So if my assumptions are wrong the results will be off.
Given what you know about my process, (right now I'm fully invested) what do I do differently when I close my next position? What would your process be?
Unfortunately the spreadsheet doesn't like TSE Stocks. I put in a symbol (TSE:CIX) and it will give me a price but the rest of the spreadsheet calculations are populated with #NUM.
I'm not sure how Google pulls in TSX data.
It's correct that if your assumptions are wrong the model will be wrong. This is the nature of any model. If a goal of a good trader is to take defined risk, your
error in defining risk is itself a risk.
Based upon your process what I would do is:
- use the sheet to test 3-5 companies in your short list -- how would they fit in with the rest of your portfolio? the biggest questions revolve around how new positions impact your portfolio volatility (standard deviation) and factor exposure
- you only want to add things that will improve your factor exposure and/or reduce portfolio volatility
- the expected return function is extremely basic, but use it as a benchmark for the return you make off the stock
note:
- time varying nature of factors is not analyzed in this model -- however, you could integrate it via changes to the model assumptions sections (e.g. say you think momentum will do very well, this year you expect it to do about 10%)
- i hope this also teaches you that using a chart to derive an estimate is pretty much garbage, but here is how you could do it (this is very bad logically and is a known cognitive bias, but...*shrugs*)
if you think linearly about a stock and this it is following a trend, your 1-year forecast would fall within that range (can take the average -- say 3700)
to generate the ER given this linear forecast (lol), you would take 3700/currentprice-1
or 3700/3469-1 = 6.65% return
you can do the same thing to factors (QMOM, QVAL, SPY, etc.)
better ways of generating expected returns are through building blocks/fundamental analysis (not what your dad thinks fundamental analysis is lol; what
equity research analysts thinks fundamental analysis is). example would be relative valuation -- if momentum p/e is currently trading at 25x and normally trades at 27x that would represent an 8% return (from the delta in multiples). you can use this in conjunction with some type of price trend forecast, but the data suggests that the past 5 years of returns is terrible in predicting the subsequent 5 years.
