Using Monte Carlo Simulation to avoid Flaw of Averages in investment

Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
#1
How we should avoid the "flaw of averages" in investment (e.g. retirement fund).

The attached PPT slides will explain what exactly is "flaw of average".

The link below is just the demo video of how monte carlo simulation is done.

https://www.dropbox.com/s/i2n6uuomfxy69p...0World.avi

(explain what is flaw of average)
https://www.dropbox.com/s/mdtosu338jzfv6...erages.avi

For pleasure reading pls.


Attached Files
.pdf   Flaw of the Averages (VB).pdf (Size: 1,010.19 KB / Downloads: 32)
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply
#2
Gross profit = Revenue - Cost

Revenue = ASP x Volume of product

Gross profit margin (GPM) = Gross profit/revenue = 1 - Cost/revenue

Rate of change (ROC) of GPM = ROC of Revenue - ROC of Cost (linear approximation)
[ Basis: ROC of a constant = zero; ROC of X/Y = ROC X - ROC Y; ROC of XY = ROC of X - ROC Y]

ROC of revenue = ROC of ASP + ROC of Volume (linear approximation).

Hence, in summary
ROC of GPM = ROC of ASP + ROC of Volume - ROC of Cost (linear approximation)

The assertion that GPM depends ONLY on ASP and Volume is correct if and only if ROC of Cost is assumed to be ZERO.

i.e. ROC of GPM = ROC of ASP + ROC of Volume (when ROC of Cost = zero)

But, I did not assume ROC of cost to be zero.
[Basis - The inflation rate in China was recorded at 2.30 percent in July of 2014. Inflation Rate in China averaged 5.71 Percent from 1986 until 2014, reaching an all time high of 28.40 Percent in February of 1989 and a record low of -2.20 Percent in April of 1999. Inflation Rate in China is reported by the National Bureau of Statistics of China]

Hope this will clarify all the doubts once and for all on my assumptions of the financial models for sunsine. I think I have been very kind to show almost every single assumption. Hope we can "move on" on this issue.
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply
#3
(20-08-2014, 11:27 AM)Curiousparty Wrote: How we should avoid the "flaw of averages" in investment (e.g. retirement fund).

The attached PPT slides will explain what exactly is "flaw of average".

The link below is just the demo video of how monte carlo simulation is done.

https://www.dropbox.com/s/i2n6uuomfxy69p...0World.avi

(explain what is flaw of average)
https://www.dropbox.com/s/mdtosu338jzfv6...erages.avi

For pleasure reading pls.


Curiousparty
Thank you for stating that the following formula was used by you:
ROC of GPM = ROC of ASP + ROC of Volume - ROC of Cost (linear approximation)

In this formula, GPM is dependent on ASP, volume and cost. So when you set values to ASP, volume and cost, GPM is fixed.
But you in your post #294, you set value on GPM too.
There are therefore 2 values of GPM, one set by you and the other determined by the formula. How did the simulation work?
Reply
#4
You are very sharp indeed to point out that there are 2 sets of GPM Smile

My ultimate aim here is to derive the possible distribution of Profit (i.e. the entire possible spectrum) based on the following 2 considerations:-

a. historical pattern of inflation
b. historical pattern of GPM (past years' performance)

The final GPM used to compute the Gross Profit is based on the "lower" of the 2 values. (use excel sheet function to perform)

tks.
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply
#5
In monte Carlo simulations I always question how do you calculate the standard deviation. This is not a number that you can use the historical distribution to forecast for the future. It simply do not make sense
Reply
#6
Yes, I agree with you wholeheartedly, i.e. historical standard deviation might not tell us the future trend of deviation.
It is the assumptions that went into the model which is very subjective.

I asked many times for value buddies to provide me with more "expert opinions" but no one wanted to give me anything. which was why I just took the liberty to do it myself.

And of course, I know there will be many critic once I released the details of my model...

I am most willing to "learn" from all the criticisms and improve my model for future rounds of sunsine's result (provided it is still "around" at that point in time).

tks.

(20-08-2014, 10:23 PM)mrEngineer Wrote: In monte Carlo simulations I always question how do you calculate the standard deviation. This is not a number that you can use the historical distribution to forecast for the future. It simply do not make sense
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply
#7
Attached is a simple real life example showing the pitfalls of relying of deterministic computation to conclude that the NPV of the project is positive.

Happy reading Smile

Comments appreciated. tks.


Attached Files
.pdf   Example of Flaw of Ave.pdf (Size: 748.94 KB / Downloads: 17)
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply
#8
I love Monte Carlo Simulation!!!!!!!!!!!!!
Its the closest thing to predicting the futre BlushHeart
Reply
#9
Monte Carlo Simulation (i.e. probabilistic simulations) is not about "predicting" the future. It is a tool to avoid the "flaw of average".

Flaw of average - simply using "mean or average value" of input parameters to estimate output value. It conveniently assumes that the entire distribution of input values can be simply represented by a certain "average or mean" value, which is totally WRONG.

This has important implications in many areas in investment, e.g. when doing DCF for companies, when performing CBA for products sold by a company, etc
tks.
[I am not here to promote any stocks. Please always do your own research before embarking on any investment decision. I will not be liable for any of your own decisions. Your use of any information or materials is entirely at your own risk. It is your responsibility to ensure that any products, services or information meet your specific requirements. I do not produce material which meets the objectives of any specific financial and risk profile of investors.]
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)