Impressive Financial Forecasting with SVR

A couple of researchers (Li, Hu, and Hirosawa) have displayed pretty interesting findings forecasting the USD/JPY exchange rate via a Support Vector Regression (SVR) network. The file explains the details. Let's break down what they did and accuracy results from the empirical test.

Basic idea of SVR: Nonlinear mapping of data into a high dimensional feature space with a kernel function, then do a linear regression in the transformed space. It all results in a nonlinear regression in low dimensional space.

Inputs of their study:

f(t): Historical USD/JPY exchange rate

F(t): Historical USD/EUR exchange rate

N(t): Historical NIKKEI 225 stock index value

O(t): Historical oil price in USD

Predicted Value (Output): USD/JPY exchange rate at the next day (t+1), next week (t+5), and 2 weeks from now (t+10)

Prediction evaluation tools:

Sum of Square Error (SSE)

Mean Absolute Error (MAE)

SSE and MAE measure deviation between actual and forecasted values, so the smaller the better the accuracy of forecasts.I find the MAE more practical of course since it isn't squared.

Correct Up Trend (CP)

Correct Down Trend (CD)

CP and CD basically gives the probability of forecast being correct for the next-day USD/JPY direction. So higher is better.

The results compared those of conventional SVR models and SVR network:

USD/JPY	Stat	SVR	SVR N
f(t+1)	sse	47.3925	36.2098
	mae	0.1672	0.1471
	cp	76.8116	77.8468
	cd	79.4118	79.8039
f(t+5)	sse	60.5918	38.4611
	mae	0.1877	0.152
	cp	76.7635	79.668
	cd	78.3465	79.3307
f(t+10)	sse	71.5295	39.7972
	mae	0.2052	0.1544
	cp	77.453	80.167
	cd	78.937	79.9213

So we can see that this is impressive for both SVR and especially the SVR Network. Directionally it's correct close to 80% of the time and an average error of within 0.15 Yen (with respect to daily range), for all three time frame predictions.

A lot of people don't look at this stuff because it takes a lot of thinking, so they convince themselves that it "doesn't work anyway". Well I'm here to tell you that IT DOES.

80% of hedge fund employees don't know jack

Haha, what the heck is going on here?
Not sure if I've mentioned this last year, someone from one of the local big Australian banks quietly informed me that their "fund managers" have no idea what the VIX is. So totally clueless regarding volatility, yet hundreds of millions of dollars are "managed" by these people. This industry is filled with morons!

Study shows upwards of 80% of hedge fund employees never actually knew what the word “hedge” meant

“It’s just like, you know, part of the title” said one trader confidently. “Another way of saying ‘badass’ fund or ‘rockstar’ fund.”...

When told that hedging involved systematically eliminating risk, most subjects just stared ahead blankly. Others checked the time.

HSBC printing money

Whoa, what the heck?!

Barak Obama, undercover samurai

Didn't see that coming, did ya?

Was George Bush so different?

Protein Cancer Connection

T. Colin Campbell, PhD and Thomas M. Campbell II point out, in "The China Study: The Most Comprehensive Study of Nutrition Ever Conducted", a statistically significant finding that diets rich in animal (especially dairy) protein could stand as the primary cause of cancer in the western world. This of course goes against mainstream American sentiment that a diet rich in animal protein (steaks, burgers, cheese...) is not only "good" but essential.

Theory

Cancer proceeds through 3 stages: initiation, promotion and progression. Chemicals responsible for "initiation" are the carcinogens, they mutate normal cell genes into cancer-prone cells. The daughter cells are forever damaged, and it begins...

Promotion sets the stage for these DNA damaged cells to grow. However, like any organic being, they will not grow unless right conditions are met. This is where diet becomes critical, to either promote cancer growth or reverse it. Having a higher Anti-promoters intake, it slows or stops, it is a push-pull process.

e.g. in order for the carcinogen Aflatoxin to take effect, it must be metabolized by an enzyme Mixed Function Oxidase (MFO). An decrease in protein intake (from 20% to 5% in their study) "not only lowered enzyme activity, but did so very quickly."

Interesting lab tests

A research from India concurs with the above. 2 groups of rats, both administed the carcinogen Aflatoxin, then fed seperate diets, one with 20% protein, the other 5%. In time, EVERY rat in the 20% group displayed evidence of liver cancer; the 5% group: ZERO. This further suggests the diet to cancer connection, instead of simply focusing on carcinogens.

Another test found that dairy protein was probably the worst thing to have if you want to avoid prostate cancer.

Don't take my words for it, don't trust the mainstream, do your own research.

Real unemployment likely at 18%

Someone at elitetrader.com got this off Shadow Government Statistics, updated Jan. 9th 2009. Truth hurts.

Most people are not closers

The following excerpt from Neil Strauss' The Game really touches base with my personal convictions. The crowd is almost always wrong, grow some balls and stand out.

"
One of the reasons I became a writer is that, unlike starting a band, directing movies, or acting in a theatrical production, you can do it alone. Your success and failure depend entirely on yourself. I've never trusted collaborations, because most people in this world are not closers. They don't finish what they start; they don't live what they dream; they sabotage their own progress because they're afraid they won't find what they seek.
"

Yes I've read The Game. Who hasn't?

Obama and economics

Yesterday, Obama spoke about the next government intervention, AKA Rescue Package to finance entitlement programs. Some believe this would hurt the economy further and others optimistic. Does it really matter?

Too many uncertainties

Obama gives speeches where he reiterates American sentiment in very broad terms in order to create rapport with the majority, and he isn’t exactly known for specifying solutions. No explicit answers lead to speculations toward what he might support (e.g. relieving a certain African American from auto, mortgage loans and etc.), and uncertainties lead to increased volatility.

The stock market has a negative correlation to volatility measures. Remember the day America announced Obama’s win, and the market dropped like a rock immediately? That is uncertainty, volatility at work.

At the same time, if the entitlement programs do surface, financing them through debt and heavier taxes could lead to a slow inflationary decline like Japan in the 90’s. Stagflation is undesirable, but it will offer an illusion of economic expansion as inflation typically pushes market indexes into higher ground.

Obama’s motivations

Campaign contributors, big money individuals and entities usually have large influence over the politicians they finance. Since information regarding campaign backer interests will not likely be divulged publically, though Webster Tarpley has speculated, Obama’s motivations remain largely uncertain.

Who cares

It is usually a bad idea to make conclusions based on uncertainties, however educated they may seem. Whatever comes next, just roll with it and adapt. Barak Obama is not that special; whatever he does will not affect the markets more than past politicians.

We have always, and probably should work with what we DO know. After all, everything we have achieved so far has been a result of our own decisions.

Most people are wrong about most things, most of the time. They're wrong about Obama.

It has begun.

Financial freedom from China

Ohh the irony...

About credit ratings

Anchoring risk management on mainstream credit ratings has resulted (and will in the future) badly for many investments. Accurate forecasts have value, conventional credit rating agencies do not. A lot of reasons point to an imbalanced system, but I think an anecdotal analysis should suffice for most of us.

Trend followers (euphemism for newbies)

Since early 07, many have raised alarms regarding the credit, housing bubbles, (e.g. Schiller, Whitney, and etc.) The credit rating agencies however remained optimistic. The mainstream investment trusts like those New Zealand finance companies relied entirely on the overly confident ratings and kept on buying. Then it started to wobble, and with each decline credit ratings slipped.

So what is the pattern here? Can the average Joe Blow provide practically identical credit ratings as those of the mainstream agencies? Yes and YES.

Here is what you do to mirror Moody’s or S&P ratings. If the market has ended lower the past quarter, lower average credit ratings, and vice versa. This is the basic concept of “trend following” or herd mentality, where you’re simply led instead of taking lead.

It is that simple, even my grandma can do it.

Victims

Many NZ based investment firms have lost everything due to over leveraging. ING credit unit trusts (Diversified Yield Fund, Regular Income Fund) hurt a lot more people with their seemingly qualified teams. These funds consist of mostly CDOs (Collateralized Debt Obligations), and even the salesmen could not explain exactly who backs the debt. Their unit prices were at roughly $1.00 June 07.

It is unfortunate that many New Zealanders became victims of this via slick selling. But then again that is the price when they chose not to do the necessary research. Hence, these investors/speculators are termed “dumb money” on Wall Street.

Bottom line

Now that we know how credit ratings work in general, you can distance yourself from the patsies, and gauge sentiment of retail investors. When credit rating for certain instruments remains high while you discover weaknesses (like some of us in 07), the extreme newbie optimism meant favorable opportunities to short. Pretty easy right?

Large short volume on US treasury

Nearly 10 million shares short, out of 15.6 million floating. Significantly large short interest on the 20+ year US treasury bond ETF means grim sentiment.

Apparently some of the bigger players are betting on the downside, perhaps from speculations of US losing the triple A credit rating, or simply large selling of these bonds due to the negative inflation adjusted return...

This is gonna be some party!

Machines running without fuel

Free energy, something generally considered impossible and "crazy". Therefore, this topic certainly deserves at least a chance to tell its story.

Something from the video worth mentioning, when you fill up your car with gas (petrol), only 30% is used for actual running fuel, the rest are lost from mainly friction and heat.

This is pretty interesting if nothing else, enjoy the video!

part 1


part 2


part 3


part 4


part 5


part 6

About the blog title change

With so many out there already writing about trading, I can probably offer more value to the readers by expanding the content into relevant, and not necessarily economics related realms. As I've found that going against the grain remains the key to success in the financial markets, I also see this in many other daily things.

I plan to discuss stuff ranging from horse racing to free energy. Of course financial economics will remain as one of the central themes. Hopefully the readers will find the info valuable however seemingly irrelavent as they are. Will elaborate in the coming posts.