The high probability of low-probability events

Non-investing personal finance issues including insurance, credit, real estate, taxes, employment and legal issues such as trusts and wills
Post Reply
trueblueky
Posts: 1376
Joined: Tue May 27, 2014 3:50 pm

The high probability of low-probability events

Post by trueblueky » Mon Jun 04, 2018 6:27 am

https://www.thesimpledollar.com/the-hig ... ty-events/

Trent Hamm uses this to promote emergency funds: Many things could go wrong. Each is unlikely. Together, it is likely one will.

Dandy
Posts: 5465
Joined: Sun Apr 25, 2010 7:42 pm

Re: The high probability of low-probability events

Post by Dandy » Mon Jun 04, 2018 6:44 am

Interesting way of looking at risk. Recently, a town in Maryland had a 1000 year flood for the second year in a row. This year a tree limb fell on my car in my driveway causing $9500 worth of damage. I've lived in this house for 44 years. In 2008 a mortgage crisis cost me a job in a mutual fund subsidiary. My wife contracted a life threatening illness (ok now) at an earlier age than most stats said would occur. And so far my family has been extremely lucky.

That is why people suggest having an emergency fund and why I always invested on the moderate/conservative side. The idea is that there are so many long shots that one or more might occur is an interesting way of looking at risk. I believe that some people have been struck by lightening more than once and that some have won the lottery more than once.

JoeRetire
Posts: 1686
Joined: Tue Jan 16, 2018 2:44 pm

Re: The high probability of low-probability events

Post by JoeRetire » Mon Jun 04, 2018 6:46 am

trueblueky wrote:
Mon Jun 04, 2018 6:27 am
Trent Hamm uses this to promote emergency funds: Many things could go wrong. Each is unlikely. Together, it is likely one will.
Stuff happens.

whiskeyChaser
Posts: 4
Joined: Wed May 16, 2018 1:21 pm

Re: The high probability of low-probability events

Post by whiskeyChaser » Mon Jun 04, 2018 7:08 am

That's why in Texas Hold'em you thin the field before the flop when holding Aces. Against any ONE set of cards, you're a favorite.

But if there are too many hands in the round, you won't be the favorite against ALL of them at the Same Time. (or, there is a high probability that one of the "low winning percentage" hands will indeed . . . win).

Valuethinker
Posts: 36640
Joined: Fri May 11, 2007 11:07 am

Re: The high probability of low-probability events

Post by Valuethinker » Mon Jun 04, 2018 9:12 am

Dandy wrote:
Mon Jun 04, 2018 6:44 am
Interesting way of looking at risk. Recently, a town in Maryland had a 1000 year flood for the second year in a row.
The way that probability was estimated was almost certainly wrong. That the first flood happened was a warning it would come again. The draws awarded to the States for the Colorado River were taken at a time which proved to have unusually high precipitation-- thus the total allowed draws exceed the river's flow, now. (And unfortunately the draws were done in acre-feet, not in percentage or share terms).
This year a tree limb fell on my car in my driveway causing $9500 worth of damage. I've lived in this house for 44 years.
Again, trees get older and more likely to drop limbs. So the probability of that event was not constant but was rising over time (and maybe not with a smooth probability increase each year).
In 2008 a mortgage crisis cost me a job in a mutual fund subsidiary. My wife contracted a life threatening illness (ok now) at an earlier age than most stats said would occur. And so far my family has been extremely lucky.
Extreme events happen in financial markets *much* more often than the normal/ Gaussian distribution would predict.

Illness I have no idea, but it sounds like you just have bad luck. My father died in an accident.

And, as you note, we are both really lucky.
That is why people suggest having an emergency fund and why I always invested on the moderate/conservative side. The idea is that there are so many long shots that one or more might occur is an interesting way of looking at risk. I believe that some people have been struck by lightening more than once and that some have won the lottery more than once.
Yes to both examples. Generally your investment approach is a prudent one.

The hard part is things like disability, where it is much more common than most of us realize (think back pain and how common that is) and it is quite hard to insure-- impossible, in fact, to fully insure against income loss (you can usually only manage 60-65%).

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: The high probability of low-probability events

Post by dbr » Mon Jun 04, 2018 9:18 am

Isn't this fact one of the reasons it is wise to carry insurance of various types? It isn't just emergency funds. There is also any range of precautions and anticipations that people practice every day. Then sometimes bad things just happen.

Maybe it is a lifestyle. Some people always leave early for something they have to do and others simply can't stand any possibility of having to wait or just can't organize to anticipate events.

AlphaLess
Posts: 1045
Joined: Fri Sep 29, 2017 11:38 pm
Location: Kentucky

Re: The high probability of low-probability events

Post by AlphaLess » Mon Jun 04, 2018 9:47 am

Dandy wrote:
Mon Jun 04, 2018 6:44 am
Interesting way of looking at risk. Recently, a town in Maryland had a 1000 year flood for the second year in a row.
This is a slightly different situation.

Pr(Flood) = low
Pr(Flood_This_year|Flood_Last_Year) is much higher.

Even low probability events have underlying dynamics. When underlying factors are 'ON', the probability of the event goes up much higher.

I think the paper is arguing about something like this. The following events are low probability:
A
B
C
...
Z

So, Pr( nu ) = low, where nu is in {A..Z}.

However, Pr ( A or B or C or ... Z) essentially equals Sum Pr (nu), nu from A to Z, which is like 26 * nu.
"You can get more with a kind word and a gun than with just a kind word." George Washington

Post Reply