Tip:
Highlight text to annotate it
X
Hello, Internet and precious metals enthusiasts! My name is John Wessel. Today is April 26,
2013 for all you future people out there. For the last couple of weeks, there's been
a pretty solid beatdown in the gold and silver price, which follows about 2 years of sideways
chop and overall trend down.
I'm very interested in money, money itself, not just having money to get nice things.
That's great and I wish I had more of it but I mean history of money, psychology of money,
science of money, etc. I'm also very skeptical of what central banks are doing to currencies
around the world. I don't think it's going to end well.
So, while I'm interested in various YouTube personalities, King World News personalities,
etc. from kind of a psychological perspective, I don't find very much good investment information
in those places. So, I just wanted to share a simple technique for mostly judging yourself
but maybe, maybe that's uncomfortable. You could use it to effectively judge bloggers
you read for example who are in the habit of making predictions.
So, what I've done here is I've actually created through Google Drive a spreadsheet. You can
use Excel or whatever. This spreadsheet's purpose is to keep track of my predictions
and see how accurate I really am. So, all I do is write the date I make a prediction
and I have to give a date by which I think my prediction is going to come true. So, I
have predicted that this will happen by June 30. My prediction is that, "Premiums on widely
available silver bullion products will return to normal at major bullion dealers." The reason
I've picked this prediction is because right now, there's a meme going around about the
physical price of silver having separated from the paper price of silver. At the moment,
there's some truth to that. If you look at like American Silver Eagles or junk bags,
for example, they have high premiums, higher than normal. Silver Eagles have a premium
of I think around $5, maybe even over $5 at some places, when normally I'd expect maybe
$3 or so, maybe $2. Then, especially, junk silver. That has a high premium when usually
that trades very close, somewhere $0 - $1 premium.
So, there right now is a little bit of a separation. So, a lot of people are hyping that up; I
think way too much. I think there's, kind of, Occam's razor, easier explanations for
why premiums might be a little bit higher now that don't involve grand conspiracies.
So, I've decided to go on the record so everyone can see how hilariously wrong I am in my accuracy
of making predictions. One of the things you find out if you start recording your predictions
is how little you really know. So! You have to state the prediction in a way that on a
certain date there's a very specific way that your prediction could be proved wrong. Right?
You have to have falsifiability.
The next thing that I do is I associate a confidence level: a probability between 0
and 1 of whether I think my prediction's going to come true. Now, here's the thing. If I
just go around predicting I'm only 50% sure - well if I'm actually right on my 50% confidence
level predictions like 70% or 80% of the time it means I'm not actually a very good predictor
because I'm right more often than I predict I'm going to be right. So, even if your predictions
come true, that's not good enough. Similarly, if you're making a series of predictions that
you're giving a 70% confidence to and about 30% of time you're wrong, you're actually
doing a very good job predicting. So, it's very good to jot down what your confidence
is.
When the day comes around, I'll go to this outcome field and I'll either put in TRUE
or I'll put in FALSE. So, I've made a little formula here if you want to look at that (=IF(E2="",,IF(E2,1.0-D2,-D2))),
if you want to make your own spreadsheet. All this does if the outcome comes true it
takes 1 minus my confidence level and it gives me that many points. So, here I had a confidence
of 98% so if it comes true it gives me 2 points. On the other hand, if the outcome comes out
false, then it's going to take the negative of my confidence level. So, I kind of lose
that many points. If I'm a good predictor then over time, when you add all these up,
it's going to come out to around zero because the times that I'm wrong are going to subtract
off the times that I'm right and if I've been right and wrong in line with my confidence
levels then those values are going to cancel out.
Now, my guess - I should almost add another prediction to this - is that over time my
number's going to get very negative. I think I'm going to wind up predicting a lot of things
with a [too high] confidence level and it's just not going to happen because predicting
specific things by specific dates is really incredibly hard. You should not be surprised
when even very smart people can't predict the future. I mean if they could predict the
future they'd be much richer than maybe just mildly successful that most smart people tend
to be.
I'll put a link to this. This is shared so you can check in on me from time to time and
see how hilariously wrong my predictions are going. But nevertheless, hey! I'm putting
myself on the record. I would encourage you to put yourself on the record. If you've been
losing money on your investments you owe it to yourself to come up with a very simple
metric of how good your predictions are. Or, if you listen to people, go through their
blog posts and every time they make a prediction in one of their blog posts, put it on your
spreadsheet. Give it an estimate of what their confidence level is. Maybe they say it or
maybe you just come up with a rule like when they say, "maybe" you put a 50% confidence
level and when they say, "This is going to assuredly happen." you give it like a 99%
confidence. Just see how damn hard it is to predict. Maybe this will teach you in a very
mathematical, or exemplary way, why you shouldn't use leverage when you invest. Because, it's
just so hard to predict the future. Thank you.