Educate: How & how often do you use decision trees in your investment decisions?

[undervalued]

Can someone describe their experiences about using decision trees when making their investments and how often do you use it? A link to an example also works or a Coursera course or text book or link to a college course, etc.

 

Thank you in advance

[StubbleJumper]

Too complicated.  Borrowing from Monish Pabrai, there's really a very simple decision tree:

 

Heads I win,

Tails I don't lose very much.

[SharperDingaan]

We do a very simple decision tree every quarter, on everything we own, over a 6 month horizon.

Up, down, no change with a simple P(x) to each; no values, and trend over time.

 

It makes you more aware of the volatility associated with your hold decision, and sets you up for hedging.

See a significant decline since the last quarter? maybe its time to lighten up by 50%.

 

SD 

[InspireByReason]

We do a very simple decision tree every quarter, on everything we own, over a 6 month horizon.

Up, down, no change with a simple P(x) to each; no values, and trend over time.

 

It makes you more aware of the volatility associated with your hold decision, and sets you up for hedging.

See a significant decline since the last quarter? maybe its time to lighten up by 50%.

 

SD

 

This is such a strange practice. Why would you sell something at a lower price?

[SharperDingaan]

We do a very simple decision tree every quarter, on everything we own, over a 6 month horizon.

Up, down, no change with a simple P(x) to each; no values, and trend over time.

 

It makes you more aware of the volatility associated with your hold decision, and sets you up for hedging.

See a significant decline since the last quarter? maybe its time to lighten up by 50%.

 

SD

 

This is such a strange practice. Why would you sell something at a lower price?

 

Last quarter I assess XYZ's prospects 6 months out as 30% higher, 50% no change, and 20% lower; net impact (30% positive - 20% negative) is +10% higher. This quarter I think its 40% higher, 40% no change, and 20% lower; net impact is +20% higher, trending +10%: hold.

 

Instead I assess at 20% higher, 40% no change, and 40% lower; net impact is -20% lower, trending -30%; sell 50% & repurchase later (hedge). 3 months from now we reassess again; the net impact is now -10% lower, trending +10%; buy in the short. If we're right, we have our original position back + a short gain.

 

As measurement periods are not fixed, & you need P(x) judgements;

you pretty much have to do it in your head as well.

 

SD