Blessed by the Potato

I just finished reading The Misbehavior of Markets by Benoit Mandelbrot.

It’s a concept book: it’s not going to tell you how to better manage your money or anything of the sort. There aren’t a lot of concrete examples, either. However, it discusses a number of important concepts, especially the important ways that we model the markets differ from reality. Models are of course important for being able to manipulate variables, to try to forecast the future, or to simplify things to get a better understanding. However, relying too much on models that are not accurate can bite us in the ass. “Clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line.”

First, a quick bit in the way of introduction for those that don’t have a strong stats background. The Gaussian, or normal distribution is one that appears in a lot of places in nature (so much so that we call it the “normal” one!), even if you’re not very familiar with it, you must have seen the familiar bell-curve shape:

It looks a little different for every group of things you choose to measure. For instance, the lengths of pencils in a fresh box may have a very “sharp” bell curve, since each one is very nearly identical in length, varying by just fractions of a millimetre. The heights of adults in the population will be a bit wider, with some people towering head and shoulders above others. The heights of kids in a school may have a wider one yet, with some being young, some passing through their growth spurt. But for all these disparate measurements, you can describe the way their bell curve looks with just two numbers: the mean, and the standard deviation (σ). The mean tells you where the centre of the curve is, and the standard deviation how wide it is.

A lot of work has gone into investigating and characterizing these distributions. A very large part of the whole field of statistics is based on manipulating the properties of the normal distribution. There are countless tools out there to help you simulate or analyze data based on normal distributions. In addition to being common, the math is also fairly well-behaved, lending itself to analytical as well as numerical approaches.

So it may be no surprise then that a great number of tools exist for financial matters based on gaussian distributions. For example, if you go to a retirement planner, and they have a fancy program that will show you the various possible outcomes for how much money you’ll have to retire on based on how much you save, saying things like “95% of the time, you’ll have at least $X available per year until you turn 100” — those software tools are based on simulating possible market outcomes with this math. Options traders have formulas to tell them how much their derivatives should be worth based on these formulas. Etc.

Yet, as Mandelbrot eloquently shows in the book, markets do not follow a Gaussian distribution. There are far too many wild swings than should be seen in a normal distribution. These swings, such as the 1929 stock market crash that kicked off the Great Depression, the 1987 Black Monday crash, the Tech Wreck, the 2008/9 global financial crisis — they simply shouldn’t happen if the markets were following a random walk with the volatility implied by a normal distribution. The standard models say that these types of events should be so rare as to be essentially impossible, like finding a human closing on the 10′ mark for height — basically, never in the course of human history should any of these events have happened, but there they are, and more that I haven’t listed, all within about one human lifespan. This volatility is, unfortunately, an inescapable feature of the markets.

Other core assumptions of the standard theories are also not true. For example, that price movements are independent. But, we know that volatility clusters. This is even something seen in nature, away from the madness of human market psychology: if you plot out the size of a river’s spring flood each year for hundreds of years, it will trace a nice bell-curve. If, using the information of that bell curve you were to make a dam so that, 99 years out of the 100 you expect the dam to last, the reservoir would be big enough to allow enough water to flow downstream to provide a water supply equal to the average spring flood, you’d find your dam would be too small. That’s because if the volatility clusters, it throws the calculations off. And that’s exactly what happens when you find that one dry year follows another follows another, while the wet ones cluster up together too — more often than should be happening if each was truly an independent event, as with games of chance in a casino.

Forgetting this assumption is one of the many factors that lead to the subprime crisis in the states. They had fairly decent models telling them how likely it was that someone with bad credit they were giving a mortgage to would default. These financial warlocks could then figure out how to tranche the loans out in a big securitization package so that the top tranche would not experience any defaults (and then later what interest rate they’d need to offer to get people to take a chance on the lower tranches). However, what they didn’t account for is the clustering of volatility — when the loans went bad, they all went bad at the same time.

And the third important way that real markets differ from our models of markets is that prices are not continuous. Many real physical things are continuous. It’s not possible for me to teleport my keyboard from the top of my desk to the floor. It must — however briefly — occupy each point along the height axis as it falls. Much of our mathematics is based on continuous functions. Markets, however, are not real, physical things. They are not continuous. However, the models assume that they are. For most retail investors, this is one of the less devastating distinctions from accepted theory: so what if prices take discrete jumps? For the fancy traders trying to limit risk though, non-continuity can break some of their techniques. One common one is the stop-loss order. You set a point at which you will not tolerate further drops in a stock’s price. If the stock is falling and hits your stop-loss price, it automatically gets sold by your broker. This can help you limit your downside.

But look at Manulife today. Let’s say a trader owned MFC and had a stop-loss set at $15.50, to limit their downside. This morning the results came out, and they were bad. The stock, which closed yesterday at $16 opened for trading today at $14.87. At no point was there ever an opportunity to sell it in the $15 range.

So, before I get off ranting about the concepts themselves, I’ll say that the book was very approachable, even if you don’t have much of a statistical bent. Mandelbrot uses graphs well to get his point across, and tries to keep things non-technical (and there is next to no math). However, it is a concept book — it’s not going to tell you how to build a better portfolio or manage risk better.

Now, having just spent some time discussing the idea that markets are more volatile than we thought, and that they are not efficient, what do I have to say?

Well, despite the volatility, for the long run the stock market has still been basically the best place to keep your money. In the short run, the volatility can be very painful (as we’ve seen recently). But, if you have many decades to wait, and believe that businesses will continue to be profitable, then you should get your money to work for you in the market. So, as someone still fairly young (despite my looks as science has ravaged my body), I have a very high equity allocation. However, despite agreeing in theory with the concept of lifecycle investing and using leverage while you’re young, I’ve never been comfortable with the idea in practice, and have never been very leveraged (and not at all at the moment).

Efficient market theory says that market participants are rational, have (roughly) equal access to information, and that information about past price changes does not enable you to predict future price changes. (The esoteric proofs of some of this rely on the normal distribution). This leads one to conclude that investing in index funds is the way to go, since beating the market is impossible. Now, I don’t believe in full efficient market theory. In particular, people are not rational, and all information is not widely and universally shared (or understood!). That’s why part of my investments are “actively” managed: I invest in particular companies I think have a good chance of beating the overall market. However that said, I do believe in a weaker form of efficient markets, which is that the average person can’t beat the market net of fees, so although I talk about my stock tips and analysis here on the blog for discussion and feedback, the only recommendation I ever give people who ask is to invest in a low-cost index fund, such as TD’s e-series funds or an iShares ETF (and part of my investments are in e-series funds).

These Rogers bills are killing us. Ok, financially, we can afford $35/mo, but it just seems so silly to spend that much (now with HST!) for basic, analog cable. We don’t see the value in the higher TV packages, and especially not in digital. Especially now that there are so many shows offered on demand online (e.g., Jon Stewart on the Comedy Network, all of BNN’s programming). I haven’t watched TV in like 2 years now, and don’t have any kind of cable at my place in London. Wayfare, unfortunately, stipulates that she must be able to put something on in the background while she veges in the living room, and hasn’t wanted to give up cable.

For a while I was using my Xbox with TVersity to stream shows from the computer, but that was a little cumbersome because it meant the computer had to be up-and-running with TVersity going, etc. Wayfare wants something simpler.

So I was encouraged to see the Iomega ScreenPlay Plus HD Multimedia Player come up as an option. It’s cheap: $150, which includes a 1 TB hard drive, and promises to be easy to use, plugging right into the TV and playing most of the common digital video formats.

First up, right out of the box it looks well put-together: there are a few pieces of foam to shockproof the drive, and the cables are in baggies, but other than that the box is split into 2 neat compartments without a lot of wasted packaging. Best of all, none of that ridiculous hard-shell packaging crap that drives me nuts. It has a remote control for operating the player part of the system, and yes, it comes with batteries. It also comes with a (rather short) USB cable, and composite and analog AV cables. Nice.

I can hook it up to my computer just like any other external USB drive, and load it up with files.

Then, it also has another USB port on the front so you can plug in another USB drive and just use the player function: something I already know will come in handy since Wayfare was a little hesitant about downloading shows and then having to unhook the drive from the TV to load them on it. She can just download them to a key or other drive, and play from there.

The operation with the remote and viewing content on the TV is straightforward. Unfortunately, the file list is as “large tiles” where the tiles are unhelpful icons just showing that you are looking at a video file. Lots of wasted screen real estate there. Only 7-8 characters of the filename show up under the large icons, with a fuller name at the bottom. Weirdly enough, even though the name at the bottom only takes up half the screen, it still truncates after 30 characters or so. I tested a few DIVX files and they all played fine — some looked great, but others had a fair bit of graininess/artifacts to them. Not sure why the difference, but those same files I’m pretty sure showed up fine on the Xbox/TVersity combo. I may have to investigate the video quality a bit more… One file that absolutely refused to play over TVersity played the video, but not the audio, so it doesn’t look like there’s any more capability there. OGG and MKV files don’t even appear in the file list as options, just AVIs and WMVs.

With composite and HDMI outputs it does have high def capabilities, but I didn’t test those out.

So from a video player point of view, it does work out of the box, but looks like it could use a few more refinements yet. Oh, one other weird bug is that it sorts capital and lower-case letters separately (i.e.: it would order a list of shows like Alpha Delta Echo November Zulu alpha bravo delta… etc.), which is just not right.

Wow, what a week for shipping.

First up, as you all know now, my preorder for SC2 didn’t arrive on time (and then the copy I picked up to tide me over wouldn’t activate!). That was with Canada Post.

Two weeks ago, I shipped via FedEx a gift for my mom. It was guaranteed to get there the next day by 5pm, and it didn’t. So, I started the process of getting a refund with FedEx’s money back guarantee. Things on the FedEx end were going fairly well, until they asked for my tracking number. I had dropped the package off at PostNet here in London, a retail store that will send out packages for both FedEx and Purolator, and never got a tracking number at the time. They said they’d email me one later when it went out, but they never did that either. The helpful rep at FedEx tried to look up the package for me in their system, and couldn’t find it. Finally, we found that it was in there with PostNet listed as the sender, not me. We got the ball rolling on the refund process, but they said I’d have to call back after the package did arrive so they knew whether to refund it as a lost or late package.

This week, I called FedEx back to finish that process off, and they tell me the refund’s already been issued. “What?” I say, “There’s nothing on my credit card.” You’ve probably guessed this next part: they issued the refund to PostNet. So now tomorrow I’ve got to schelp on down there and try to get them to forward the refund on to me. Any guesses as to whether or not that’s going to be an easy process, if I can get it done at all?!

Today, I get a call from the Purolator robot: there’s a package for me to pick up at their depot by the airport. First up, I hate the stupid Purolator depot by the airport. It is so ridiculously far from the rest of the city. The busses don’t go anywhere near there, so I don’t know how they expect people without a car to get their packages, and even with a car it’s a 10 minute drive each way to the very edge of the city. Second, they never even tried to deliver it to my house. I was home all day waiting for my StarCraft delivery, so I know the Purolator guy never came either, and there was no “missed delivery” note on the door.

Anyhow, once I get there, this really rude lady tells me I can’t get my package with just my ID & address, I need the tracking number. Their computer system can’t look up by name or address! So I have to phone myself to pick up the message — several times to get all the digits of the tracking number down. During this time 3 other people come in and all get the same treatment — no tracking number, no package. One of them, having driven all the way out to the boonies by the airport with his ID in hand, is sent home to go look up the tracking number because they don’t have internet there for him to check his email. Then I get the package and see a note pasted on in confirming my suspicions: “construction on street, did not attempt delivery”. Seriously, there is construction on my street, a block away, but come on, it’s summer. Construction happens, and it’s not like it was anywhere near blocking the ability to get to the house, or even park the truck on the street here!

Just got back from seeing JoCo and Paul & Storm in concert. It was a really fun concert, including at least 3 separate songs about mad scientists who are sad, so it was like they were singing right to me. Paul & Storm have a fairly large catalogue of stuff on their site, and I find most of it to be hit-or-miss, but in person (where it’s really just the hits) they were great.

It was an intimate setting with lots of feedback (often hilarious) from the audience of hard-core fans (and really, there are really only three kinds of people: those who have never heard of JoCo (yet), those who are already die-hard fans, and those who aren’t the slightest bit geeky and/or without any sense of humour and thus have heard of JoCo/P&S but aren’t fans). However, the seats in the Enwave theatre were horrendous, they weren’t deep enough to sit fully back on (it’s like when the monster truck radio guy says “you get the whole seat, but you’ll only need the edge” — well they only gave us the edge). They also were connected down the row, so whenever someone would move your seat would shake and twist. At least the venue had good acoustics.

Verdict:
Jonathan Coulton – Awesome.
Paul & Storm – Hella Awesome.
Enwave Seats – NOT awesome.

The funniest part of the concert though was Wayfare. During the Mandelbrot Set she just about lost it laughing (and when I first played the song for her, she laughed for like 15 minutes at the “best line ever in a song”). I’m pretty sure it carried right to the front, but JoCo didn’t seem to mind. Then during the encore, all 3 of them come back out to play a cover of TMBG’s Constantinople, and Paul brings out this weird mouth organ blowy thing, and JoCo and Storm comment on it and make a few jokes. Wayfare grabs my arm, turns to me, and with this really intense expression on her face says “Have you ever seen my mouth organ blowy thing? It’s red.” And partly because of how important she seemed to think it was that it was red, and mostly because of how intensely she said it, I just lost it and started laughing like crazy.

Finally, it was really weird to see so many people in “Skullcrusher Mountain” T-shirts (including me).

Just a quick note: there will be minor spoilers all through this.

I was trying to think of how to review Iron Man 2 when Orson Scott Card beat me to it.

It was a pretty movie that was fun to watch. But there were a lot of holes in it, especially with the characters and their motivations.

For Tony, he went through this whole nonsense of discovering a new element to upgrade his arc reactor to get around a problem with heavy metal poisoning. Supposedly, his use of the Iron Man suit exacerbated the leeching of toxins into his body. However, this made no sense to me, since he obviously managed to make power suits with their own power supply (such as the one Rhodie took), and only needed the power from a car battery to power the electromagnet in his chest. He easily could have externalized the arc reactor and used a conventional battery to keep him from dying.

I also didn’t like Don Cheadle as Rhodie, he just played it so straight. And as Tony’s friend, it made no sense to me why he’d steal the armour (even if under orders) just to give Stark’s competitor, Hammer, access to it. In all his interactions he was just such a divergence from what we saw in the first movie, where they skirt that line between professional camaraderie (or even babysitting) and drinking buddy.

I also didn’t think we got a very good look at Vanko — it wasn’t quite as bad as Star Trek for glossing over the villain, but Jebidaiah from the first one was just so much more real and sinister, and we got to see him develop and go power-mad before our eyes…

Anyhow, for all its shortcoming it was action-packed popcorn-munching fun, so go have at it!