Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Yesterday I worked all day on organizing the relationships apriori found on my database into a complex nested if statement. Then I made that into a new attribute.
It turns out J48 found just as many "hits" (correctly classified class values) .
That was initially discouraging except when I realized the relationships (rules) guided me in reducing the attributes to just the key ones. It also gave me a better roadmap of what which formulae are predictive and which are repeating the same thing.
It's a huge amount of work and I have been quite tired lately.
Nevertheless it shines a light in the tunnel.
Result of this step:
65% correctly classified on
up day 210 of 260
down day 273 of 324
other (inside and outside days) 0 of 164
I have been working today on the middle instance, day type "b".
There are:
260 | a = -1
165 | b = 0
330 | c = 1
I have just had a thought that the ML aims for the two most numerous groups a & c.
For b of which there is only 165, it just splits them into a or c.
I'm going to try eliminating records in category a and in category c so they are less numerous than b and see if can then be better at identifying b instances.
There can be no intelligent action coming from a negative vibration -EVER!
IF energy is in a negative vibration it must move into a negative form.
Min 19
Source: Bob Proctor - The Greatest Power In All of Creation
I was reading a trading journal which started off with the assertion that:
1. There can be no system that can predict and all movement is random.
and then later
2. You need at least a 65% edge to breakeven
Earlier in the video Bob says:
"If an idea is put forward don't ask is it right or wrong - that's not the question.
Rather ask yourself will this idea take me in the direction I want to go?"
If it won't reject the idea because if it is not taking you where you want to go it will take you where you don't want to go.
If you entertain #1 and #2 (remember now we are now debating if they are right or wrong but rather will they take me in the direction I want to go?) then why are you trading?
One's egotic mind can get caught up in the love of debating if something is right or wrong and in so doing pull the person into a disabling negative vibration and create negative results.
Here is the target or goal:
=== Confusion Matrix ===
a b c <-- classified as
0 0 260 | a = -1 => down days
0 0 164 | b = 0 => other days (read inside days and outside days)
0 0 328 | c = 1 => up days
(use classify, rules, ZeroR to do this)
If later when running your classifier if you have all the values on the topleft-bottomright diagonal you have 100% correctly classified.
I wanted to do a better job of classifying the other days.
---------------------------
In my spreadsheet I sorted days for just those that were "other" then looked for a rule with the Apriori
(use associate, Apriori)
Finding some higher percentage associations I made and If then rule in my spreadsheet to possibly identify 48 of the 164 other days which would be down days.
I used this rule to modify the values of one of my attributes
Then I ran j48
(use classify, trees, J48)
and I now have 77.9%
=== Confusion Matrix ===
a b c <-- classified as
248 2 10 | a = -1
60 54 50 | b = 0
33 11 284 | c = 1
The year was 1720. And one of the smartest people to have ever lived had just made one of the dumbest financial mistakes imaginable.
Sir Isaac Newton was a genius in every sense of the word.
He practically invented the science and mathematics that is at the foundation of nearly every bit of modern technology that we enjoy today.
Newton was such an intellectual superhero that even Albert Einstein idolized him.
In fact Einstein wrote in a 1919 paper that “[Newton’s] clear and wide ideas will forever retain their significance as the foundation on which our modern conceptions of physics have been built.”
Yet Newton was a complete moron when it came to investing.
During his lifetime, the British Empire was becoming a major superpower and had colonies all over the world.
With so much new international trade under its control, Britain’s prosperity soared.
A handful of companies like the East India Company provided opportunities for investors to share in that prosperity. But the public was always clamoring for more.
So in the early 1700s, the British government chartered a new company– the South Sea Company– and awarded it a total monopoly on British trade in South America.
It seemed like a veritable goldmine, and investors clamored to buy shares.
Isaac Newton was one of those investors.
And initially it was a fantastic investment; Newton bought in early 1720, and within a few months he’d doubled his money. So he sold his entire stake.
Then something interesting happened.
The South Sea Company’s share price kept climbing… higher and higher.
In fact, almost right after Newton sold out, the South Sea Company’s share price climbed exponentially, reaching a peak of nearly GBP 1,000 by mid-1720.
(That would be worth nearly $300,000 today.)
Newton felt like a total buffoon for sitting on the sidelines while all of his colleagues were still makings tons of money in the stock.
So he got back in.
And, anxious to make back the profits he’d missed out on, Newton doubled down, investing an even bigger amount in the shares.
You know what happens next–
The South Sea Company turned out to be a complete bust. It turns out that Britain never really developed much trade with South America.
Yet the company had blown through most of the money, and there was nothing left for the shareholders. So the stock price quickly crashed.
Newton was broke. He lost his life savings, just seven years before his death.
Now, I’ve told this story a few times in this letter… because it’s so powerful.
Even one of the smartest people who ever lived made a terrible and completely avoidable mistake because he was driven by emotion instead of reason.
However, as we explained last December, this is a low-ball estimate which "understates the potential losses" as it "does not include high-yield bonds, fixed-rate mortgages, and fixed-income derivatives", which would suggest that the real number is likely more than double the estimated when taking into account all duration products. As a reminder, Goldman calculated the entire duration universe at $40 trillion as of the summer of 2016, resulting in $2.4 trillion in losses for a 1% move.
Speaking at Davos, Dalio also predicted that the Federal Reserve will tighten monetary policy more than they have signaled, and said that "economic growth is in the late stage of the cycle but could continue to improve for another two years."
Echoing what he said on CNBC yesterday, Dalio said that The current economic environment is good for stocks but bad for bond investors. “It feels stupid to own cash in this kind of environment. It’s going to be great for earnings and great for stimulation of growth."
In contrast to Dalio's public talk bravado (who may well be "taking his book") the current environment could be great for an equity crash.
======
What investors with bonds would do in the case of a large drop?
Of course they could sell bonds - but another alternative is to sell the portion of their portfolio showing gains - equities.
Usually people sell some of their winners.
If the dollar is falling then foreign bond holders will get hit twice - once on the falling price and again on the currency exchange.
After all, rising interest rates can cause a large increase in credit defaults - mortgages, lines of credit, auto loans etc- triggering job losses and more defaults. Those individuals with equities are likely to sell some to cover bills and mortgages. If selling picks up others will sell so as to not go through 2007-9 again.
Those individuals without equities and put out-of-work will need to govt assistance. Higher rates mean higher cost of services govt debt. So govt costs will rise from two sources.