John St. John's blog

Do the mutual fund recommendations made in one of the most expensive investment books of all time hold up?

noreply@blogger.com (John St. John) — Sat, 04 Jan 2014 18:43:00 +0000

Does the mutual fund advice in one of the most expensive investment books hold up to the test of time? As of me writing this introduction paragraph I have no idea what the answer is to this question.

Seth Klarman, author of a well received currently out of print investment book that sells on Amazon for nearly $2,000 (see http://amzn.com/0887305105), lays out some investment advice to those of us who don't invest full time, in the last chapter. He recommends either going with a value focused mutual fund, or finding a talented and dedicated investment advisor who is on board with value investment fundamentals. For those of us with less than maybe $100-250K in assets to hand over to an advisor, it will be difficult to find someone who can give the kind of dedication that Klarman recommends looking for in an advisor, leaving a mutual fund as the only viable recommendation.

Interestingly Klarman lays out a few specific recommendations for value focused mutual funds that were available at the time in 1990. Since 1990 we have gone through just about every market condition, including one of the most major crashes in the history of the stock market, along with a few super impressive bubbles such as the dot com boom of the 2000's.

Of any segment in time to really test how an investment philosophy pans out relative to something else, between 1990 and today seems like just about the best test scenario you could hope for.

So the next question is what do we test Klarman's recommended mutual funds against? Well Klarman really pans the idea of index funds. In fact he incorrectly (as of today) predicted that index funds would go the way of the dinosaurs. However when you consider the growing popularity of index ETF's, along with Vanguard, one of the largest mutual fund companies in existence making its bread and butter predominately on cheaply managed index funds, this can't seem farther from the truth. Perhaps this prediction will pan out in the future, but it is not looking so good as of now. Taking this into consideration, and his mention of the S&P 500 as an example of an index that people like to invest in, lets compare Vanguard's S&P 500 mutual fund returns to Klarman's recommended value funds over the longest time-frame possible.

Klarman's Recommended Value Funds vs S&P 500 Index Fund Bake-off

So Klarman recommends two families of mutual funds, and recommends against class A shares with their front-loaded fees. The first is Franklin Templeton's Mutual Series Funds of which the following were available near the beginning of 1990 (FCMRX, TEMEX, TEMQX, TEMTX), and the Sequoia Fund (SEQUX). The vanguard fund I am comparing is the S&P 500 investor index fund (VFINX). There is currently a cheaper "Admiral" version of the Vanguard index fund, but that only became available in 2000 as far as I can tell.

The chart I put together on Morningstar's web site comparing these funds should be available at this link.

I compared two scenarios, the first would be if Klarman were to recommend that a friend invest in these funds in 1990 while he was writing this book and examining these funds. I chose 2/28/1990 because that was the earliest I could get the slider in that year. The other scenario is that in early 1992, you just read this book when it was hot off of the presses, and invested in these fund recommendations.

Klarman's value mutual fund recommendation outcomes had you invested in 1990 around when he was likely writing his book.

Klarman's value mutual fund recommendation outcomes had you invested in 1992 just after reading his book when it was hot off of the presses.

Whether you invested in 1992 just after reading this book when it was hot off of the press, or in 1990 when Klarman was likely considering available mutual fund options for this chapter, it really doesn't appear upon quick visual inspection that his recommendations did phenomenally better than what he strongly advised against: a mutual fund based on the S&P 500 index. They do appear to generally beat the index fund though, with a single exception, the Franklin Mutual Recovery C fund which performed a little worse.

Most of these value funds do appear to exhibit the telltale signs of being a value fund. For example notice the peak around 1999 that corresponds to the dot com boom and bust. These funds tended to be flatter around that peak, which Klarman said typically happens with value funds when the market gets really hot.

One of Seth's other assertions about value fund performance is that when the market as a whole drops, value funds if they indeed are investing in value stocks, have a greater margin of safety so they will tend to drop in value less. When we focus in on 2007-2008 between the peak and trough of the market, we would predict that these value funds would have lost less value than the naively chosen S&P 500 index. This is the case with some funds, but not others. Mutual Beacon, Mutual Shares, and Mutual Recovery perform worse than the S&P in this window, while the other 2 Sequoia, and Mutual Quest outperform the index.

2007 peak to end of 2008 market low point.

Another assertion is that given the value opportunities that one would get from a severe market crash like the one that happened in 2007-2008, there would be really nice value buys available to these savvy investors. However even in this fairly optimistic window, only Sequoia outperformed the S&P 500 index.

2008 low point until today's performance

Indeed it seems like the only thing that really helped these value funds in the long term generally outperform the S&P 500 index was the dot com bust of 2000 through 2002. Indeed in this window, each of these funds gained the upper hand on the index.

So what do I do with this information?

It really is not clear to me that these professionally managed mutual funds handily outperform a cheap index fund of stocks over the long term. While Seth Klarman's arguments in favor of value investing make a ton of sense to me, and people like Seth Klarman in the hedge fund world have apparently done very well over long periods of time, these value focused mutual funds recommended at the time by him did not clearly withstand the test of time. For example if you and 4 of your friends had read his book in 1992, and each of you invested in a different one of the 5 value mutual funds recommended by Klarman at the time, one of you would be fairly upset that you had not chosen to simply invest in a cheap index fund! That one person would have 27% less money than had they invested in the index. Also one of you would be pretty indifferent because you would only have a fraction of a percent more than had you invested in the index fund.

On the other hand the other three of you would be better off by between 27% and 62%.

Two of your group of five friends would be at a wash (or worse) while three of you would be quite happy with Klarman's advice, 22 years later. This reminds me of a book I read a while ago that he scoffed at in his writing, a random walk down wall street. Although there is probably some non-randomness in skilled value investor's abilities to chose good investments, when you factor in the higher investment fees to pay those professionals to do their job, along with random bad luck, there does not appear to be a free lunch here!

How to securly encrypt your emails.

noreply@blogger.com (John St. John) — Fri, 22 May 2009 02:35:00 +0000

I had trouble finding a good online tutorial walking one through all of the steps necessary to set up gpg, set up private/public keys, and use that to communicate securely with a friend. Anyhow the following is a quick tutorial of how to do this on Mac and Linux. If you are on windows try: http://www.gpg4win.org/ and see if you can use my tutorial along with whatever how-to guides they have to get it working.

If both you and your friend are new to Gnu Privacy Guard, then both of you should do the following.

1. install gnu privacy guard:
mac: http://macgpg.sourceforge.net/ and download current version
linux: it may already be installed in your system, on the command line type: gpg and see if anything happens. If not google "install gnu privacy guard [the name of your linux distro]" and there will surely be a tutorial.

On the command line:
2. type gpg --gen-key

3. I would select the first option DSA/emgammal or whatever

4. Choose the highest available encryption

5. Follow the rest of the instructions to finish generating a key. Your passphrase should be a longer sentence like "oh no i do not know how to type anymore after that accident"(please not that sentence though) and somehow remember that sentence.

6. type gpg --list-keys

7. You will have something that looks like:
pub 1024D/F217E383 2009-05-04
uid John St John (my launchpad key)
sub ******************************

****** (I blanked out my secret key info)

In the above example, the underlined text portion (F217E383) is my public key's ID. To upload that key to a keyserver so anyone can send you an encrypted file type:

gpg --send-key Your Key ID

8. To get my key so you can send me an encrypted file type in:

gpg --search-keys 'your_friend's_email_address@whatever.com'

and double check with your friend on phone or in person that the key you see is in fact theirs. You can also do the search by name, or probably key ID.

9. Now make a text file say "secret.txt" for example and type whatever message you want in it.

10. When you are in that folder on the command line type:

gpg --encrypt --sign --recipient 'your_friend's_email@whatever.com' secret.txt (or whatever your file is called)

follow the instructions...

11. you should have a file now called secret.txt.gpg that is an encrypted file that is impossible to decrypt by anyone who doesn't have access to your friend's private key.

12. add that file as an attachment to an email to your friend.

13. Have you friend download the attachment, then on the command line change to the directory that the file was downloaded to, and type:
gpg --decrypt file_name

Dividing Fractions in the Real World

noreply@blogger.com (John St. John) — Wed, 25 Mar 2009 02:22:00 +0000

Let me preface this by stating that I learned the trick for dividing fractions, simply multiplying one of the terms by the other's inverse, a long time ago. However I always find things more understandable when I have a good concrete example of how something works. My girlfriend is very strong in math, and wants to be an elementary or middle school math teacher, and she gave me the following example to think about. Say for example you have 4 quarts of ice cream, and you want to split it up into cups that each hold 2/3 of a quart. First you take 2/3 of each quart of ice cream, leaving 1/3 quarts in each ice cream container. Since you started with the 4, you have now filled up 4 cups. You have four 1/3 quart portions left and you want to fill up 2/3 quart cups with that portion, so two of those 1/3 quart portions will fill each cup, meaning two more cups are needed.

I thought it was a very good way to think about division. Perhaps I was taught an example like that in school, but all I remember was the simple inversion trick. I didn't get a really good understanding for what I was doing.

Another thing that helps with fractions is to change what I subvocalize when I see one. I was taught to think "two thirds" when I see 2/3 for example. If on the other hand I think "out of three parts, take two", when I see 2/3, I get a much better understanding for what the fraction actually means.

fonolo 4 Android: A look back at Software Engineering

noreply@blogger.com (John St. John) — Mon, 16 Mar 2009 20:51:00 +0000

This term taking Software Engineering with professor Anthony Hornof (credit him for the image to the right) was a great experience. My team developed an application for the Android phone which we subsequently distributed on the Android marketplace, and is downloadable in the US on any Android based cell phone. Learning how to develop this type of application, with the end user in mind, and also working on managing the team was definitely a valuable experience in the long run. If any future developers are interested in checking out the source code for this application, it is available at http://code.google.com/p/fonolo4android/ Perhaps I will crank out another Android based application in my free time in the near future.

My Android App

noreply@blogger.com (John St. John) — Sun, 25 Jan 2009 05:13:00 +0000

Now that I have the critical aspect of my group's Android app up and running, I wanted to say a little about how cool it could be. It is an application that integrates with an external service called fonolo, and will provide a clean Android specific user interface for working with fonolo.

What fonolo does is really cool. They eliminate the need to navigate those annoying phone menus when you call a typical large corporation. Instead they maintain a text representation of the options, and you can quickly skim through to select the exact location you want to get to. When you submit the request to fonolo they place the call on their end, automatically navigate the phone menus, and then call your phone with the company and department you want on the other end. The android app makes it even cooler because rather than having to navigate to the website and log into that, you can do it directly from the application in your Android based phone!

One big thing I am learning from this project is how cool of an operating system Android is. It is incredibly easy to write programs for the system, and you can write code on any operating system that supports Java(unlike the iPhone which requires OSX and XCode as far as I know). Additionally to deploy your application on the Android marketplace, you only have to pay a one time fee of $25 as a developer. For the same privalege you have to pay $99 for the iPhone marketplace. Once more Android based handsets come out I am sure Android will be very competitive in the internet phone marketplace.

Winter Break

noreply@blogger.com (John St. John) — Mon, 22 Dec 2008 18:50:00 +0000

I just finished a very stressful term that involved applying to grad school, and three tough computer science classes. Now that it is over, I am already getting ready for next term and a whole new set of challenges. This coming term I will be writing up and presenting my undergraduate thesis based on work that I have done on and off since September 2007 under Joe Thornton and Bryan Kolaczkowski. Additionally I will be working on Software Methodology, a demanding Computer Science course that centers around a group programming project. I already have some ideas of what I would like to work on for that course. I am getting interested in Android, Google's new open source mobile phone OS. Developing apps for it seems pretty straight forward. Also once the apps are done, marketing them is as easy as uploading them to the Android marketplace. Although Android doesn't currently reach as broad an audience as the iPhone, I believe that it will eventually overtake the iPhone in the global marketplace. Additionally while the iPhone requires programming in objective C, which I have no experience with, Android is built on Java which I am fairly comfortable with. All in all Android seems a much better investment of my time and energy.

Hope everyone is having a happy holiday season.
-John

Eukaryotic cell organelles

noreply@blogger.com (John St. John) — Tue, 30 Sep 2008 21:09:00 +0000

In preparation for the biology GRE subject test I need to re-learn (or learn) general biology. My upper division work has been focused on evolution and computational techniques, so I am very rusty on most other biological topics.

Today I went over organelle structure and function in eukaryotes (organisms that have membrane enclosed structures within their cells such as animals, plants, protists, and fungi). As well as containing the chromosomes, the nucleus contains the nucleolus which is the production site for ribosomal subunits. These subunits are then shipped out of the nucleolus, then the nucleus and into the cytoplasm. The ribosomal subunits are destined to either float freely in the cytoplasm where they eventually link up over messenger RNA (mRNA) to make proteins destined for the cytoplasm, or they attach to the walls of the endoplasmic reticulum (ER) where they produce proteins that are destined for a membrane or export out of the cell. It is also interesting to note that the mitochondria and chloroplast produce their own set of robosomes that more closely resembles procharyotic(bacteria, archea, etc) ribosomes.

The destination of proteins is determined by their amino acid sequence, structure, and sometimes post-translational modifications. Proteins that are to be secreted out of the cell have a hydrophobic signal sequence called the signal peptide.

Lysosomes are interesting. They have enzymes inside of them which digest proteins, carbohydrates and nucleic acids. These enzymes function well in an acidic environment where the pH is about 5 (a cell's pH is closer to 7 meaning it is neutral). This fact keeps the overall cell safe because the enzymes aren't able to digest the cell in its higher pH environment. However inside the lysosome, the pH is kept at 5 so that the enzymes are able to do their work. That is a very interesting adaptation.

Peroxisomes are similar and break down fats and harmful chemicals like alcohal. These organelles do so through the production and degredation of peroxide, which if allowed to come in contact with DNA would be very dangerous. Thus the reactions are kept within the peroxisome.

The mitochondria and chloroplast are cool. They have their own DNA, and produce many of their own proteins. They are basically like a small unicellular organism that lives within the cell and produces the energy needed for the cellular function. Biologists theorize that these organelles were once seperate organisms that entered into a symbiotic relationship where they exchanged energy production for protection and a stable environment.

Programming languages

noreply@blogger.com (John St. John) — Tue, 30 Sep 2008 01:43:00 +0000

We covered some of the basic programming language types today. The imperative type is the most commonly used one. It is characterized by assignments and statements following a sequential order. C and Java are examples of this. The functional type is less common and includes languages such as scheme, lisp, ML, and Haskell. In its purest form the functional language contains no assignments. The entire program is defined by function definitions, and the invocations of those functions. This idea of a language is called lambda calculus. Logic languages seem pretty interesting. The idea of a logic language is that you define logic relations between variables, and then when you implement the program, you can have it solve for any number of variables. The simple example of this we got in class was say you have a function called append. Rather than defining the function like one would in C where you take in two things and spit out the appended result, in a Logic language like Prolog you have the relational function append that has three things as arguments and the function returns true if the three things satisfy the defined requirement of the first two being subsets of the third. If you pass in a variable you want the function to solve for it will do that for you and spit out all of the solutions. For example ?append([1,2],[3,4],X) would spit out [1,2,3,4]. If you give it ?append(X, [3,4], [1,2,3,4]) it would spit out [1,2]. It seems like logic languages could, in theory, simplify programming.

Artificial Intelligence

noreply@blogger.com (John St. John) — Tue, 30 Sep 2008 01:29:00 +0000

This is the first post of what I plan on being many posts about what I learn from day to day in my classes. Yes, this may be a good time to unsubscribe from email updates.

AI is extremely prevalent. More so than I previously thought. When I think of AI I typically think of humanoid robots like Kizmet. Examples that I wouldn't have thought of include the intelligent agents in computer games, wall street trading robots and risk management systems, Google web search, the mars rover... There are tons of examples out there. In fact every computer utalizes concepts that were first thought of in AI research. For example the idea of hard coding expert information into computers was an active area of research in the 80's; today searchable dictionaries, computer diagnostic programs, and online sites like webmd are commonplace.

I am very excited to continue on with this class. I plan on posting a few things that I find the most interesting from each lecture.

Olympic Trials

noreply@blogger.com (John St. John) — Fri, 27 Jun 2008 21:15:00 +0000

The Olympic trials event started today. I don't think there have been any races or anything but the fair is going on, and a ton of people are here. I actually live across from one of the corners to Hayward field where the event is occurring so it is very exciting.

My girlfriend and I participated in one of the events that is probably supposed to get visitors to campus to check out the UO. As a result I won this cool "limited edition" medal. For getting a sheet stamped at a few places around campus. Well worth the effort if you ask me.

Summer School

noreply@blogger.com (John St. John) — Fri, 27 Jun 2008 21:13:00 +0000

I signed up for a few online classes thinking they would be easy. Unfortunately, although they are easy, they involve a lot of work! Oh well. I am learning about US Politics, International Relations, and English Linguistics.

Poster on my research & Atheists who make "moral" choices

noreply@blogger.com (John St. John) — Sat, 10 May 2008 19:51:00 +0000

I am currently making a poster that I will be presenting at an evolution and development conference in Berkeley on the 28th. I am both nervous and excited. It is going to be on work I have done and discussed briefly in this blog for the past year or so. The toughest part will be trying to make the mostly mathematically and computer science based work I am doing interesting to evolutionary biologists. It should be interesting, it is very relevant to that field, I just need to come up with a good way of presenting the data.

Last night I went to a Barack Obama rally here at the University of Oregon. It was a lot of fun and I really hope he gets the presidency. Afterwords I joined some friends from the CIS department for beers, and when they asked me what I was having I informed them that I am one of those weird Atheists who doesn't drink. Really it isn't very weird at all, it is a very logical decision especially given that I have substance abuse issues in my family and showed signs of them myself in high school. That said I had a really good time and didn't feel awkward at all. I usually get very uneasy when telling people that I don't drink because I am afraid they will label me as a religious fanatic of some sort. This time though preceding the "I don't drink" with letting them know I am also an Atheist, really helped take the awkwardness out of the situation for me. I think I will continue to handle social situations involving drinking in this way.

Phylogenetic methods talk

noreply@blogger.com (John St. John) — Thu, 06 Mar 2008 03:49:00 +0000

Today I gave a 10 min presentation on phylogenetic methods and evolutionary model selection. So I talked about how evolutionary trees are made from genetic sequence data. I think I nailed it pretty well.

After the talk, this guy asked me if you could use this stuff to predict where evolution will take a species in the future. I told him no, unfortunately to predict that you need to know what the environmental pressures will be, and there is a huge random component as well. I went on to tell him that the really interesting thing that you can do with phylogenetics is, fairly accurately, determine the genetic sequence of the last common ancestor of the species you test. The lab I am working in actually recently did that to help refute one of the main anti-evolution arguments posed by creationists. Creationists say that complex systems such as hormone-hormone receptor interactions are too specific to be explainable by evolution. That argument makes sense since hormone receptors are often very specific to the hormone that they bind. It seems that to explain the evolution of such interactions you would have to literally produce both hormone and receptor in the same lineage and it would have to immediately serve a purpose and rise to fixation. If that were true, it would be nearly impossible to do such a thing evolutionarily.

People in my lab sequenced estrogen receptors and estrogen related receptors from a very wide range of organisms to try and get at a very ancient common ancestral hormone receptor to examine it and see what it was like. They then took the sequence generated from the ancestor and grafted it onto yeast cells so they could examine which hormones it was sensitive to. They found that the ancestral hormone receptor was able to bind a variety of ligands including estrogen, which hadn't evolved yet. That showed that you in fact didn't need to evolve hormone and hormone receptor at the same time, that you could start out with a really unspecific hormone receptor that worked with completely different hormones which was eventually taken advantage of evolutionarily to develop into several different, specific hormone receptor-hormone interactions.

Evolution of Altruism

noreply@blogger.com (John St. John) — Thu, 06 Mar 2008 03:48:00 +0000

Today I learned about the evolution of altruism. Darwin admitted that altruistic behavior seemed to go against the idea of evolution. Recently though people came up with the math to show exactly how, even though it has a negative fitness to the individual who practices it, it can actually have an overall positive fitness effect. Here is the formula: rB > C where C is the negative effect to the individual for practicing altruism measured in # of offspring, B is the positive effect of altruism measured in # of extra grandchildren, and r is the proportion of genetic similarity between the altruistic individual and the individual who benefits directly from the altruism. Here is an example: ground squirrels practice altruism by giving warning cries to alert their fellow squirrels about the approach of predators. The squirrel giving the cries puts itself in danger by being an easier target for predation. However more of the offspring will survive if a squirrel does it. So here is the math, say on average a squirrel who has the gene causing it to give warning cries has a shorter life, and on average produces 2 fewer offspring, however the offspring it does produce has higher rates of survival, say 5 of them survive that would have otherwise been eaten. For a given squirrel since half of the DNA comes from mom and half from dad, the r will be equal to 0.5. doing this math we get 2.5> 2 which is true so you would hypothesize that this altruistic behavior would be genetically selected for. This is the same principle that goes into the evolution of extreme altruism as seen in ant colonies. The purest form of altruism is the worker ant who is completely sterile, it has no chance of having offspring. The reason this evolves though is because ants are so inbred that the value for r is extremely high. Any ant in the ant colony shares about 0.8 of its DNA with any other ant so you will hypothesize that any altruistic behavior will be very beneficial to the fitness of the gene that produced that altruistic behavior. Just for a comparison if you compare the entire human population the likelihood that you share genes with any other individual is about 0.0001, which is why we haven't evolved into having classes of drone workers like the ants have.

Sympatric Speciation- How to divide a population that is not physically divided.

noreply@blogger.com (John St. John) — Tue, 19 Feb 2008 08:50:00 +0000

I will share about sympatric speciation (evolution of one species into two when all populations live together in the same location) and how it ties into the process of evolution as a whole. It is a common misconception that for speciation to occur, populations need to be physically separated from each other. This need not be true, although it is correct that the process must be a result of some form of reproductive isolation there are subtle differences that lead to some methods being viable and others not. First I will define species as two populations that cannot interbreed to produce successful offspring. There are two forms of reproductive isolation, one is prezygotic isolation and the other is postzygotic isolation. Postzygotic reproductive isolation is the phenomena of two populations that are able to breed but always produce offspring that have significantly reduced fitness, and are thus unable to withstand the forces of natural selection over time. Examples of how this would look are: the outcome is lethal to the heterozygote offspring, they could come out sterile, or they may simply have reduced performance in the environment. Postzygotic reproductive isolation can only successfully evolve in populations that are physically separated from each other, this is because if a population is all together, the alleles leading to the reproductive isolation will be selected against because they result in individuals with decreased fitness and thus they will not ever rise to fixation. This is by the same process that two populations that continue to have postzygotic reproductive isolation will never successfully come together again, they are weeded out by natural selection.

Prezygotic reproductive isolation on the other hand is observed when individuals are physically incapable or preferentially driven away from mating outside of their group. These individuals are not selected against by natural selection because they do not end up cross breeding in the first place, thus they do not actually realize problems with heterozygotes having reduced fitness. Examples of this occur when different populations have different mating times so they never end up interbreeding, there may be gamete incompatibilities where the egg of one population rejects the sperm of another, and finally it could be due to the evolved mating preference of the individuals of each population. The fixation of allele frequencies that lead to prezygotic reproductive isolation usually occurs as a result of those alleles leading to an increase in fitness. One example of this in the wild is the stickleback which is a fish found in both lakes and oceans around the world. This fish is currently undergoing speciation due to prezygotic isolation. there are two emerging species of this fish, one typically lives away from the bottom of lakes, and the other is a bottom feeder. I specify that they are emerging species because even though the populations can interbreed, and sometimes do given the right environment, they have already evolved significant differences and inhabit separate niches. The one that lives away from the lakebed has spines and other things that help protect it from predation while the one that lives on the bottom doesn't have spines but does have features that allow it to feed better in that area. When the species are interbred the half-breed that is produced has less fitness than either of the two populations. Thus you can see how the development of mating preference would be advantageous to the survival of each of the sub-populations. It turns out that the probability of spawning between the bottom dwelling sticklebacks and the sticklebacks that live away from the bottom is much lower than within the populations. As a control the two groups were given the chance to breed with populations from separated lakes and it was seen that they do not mind matting with individuals from the other populations as long as they are of the same evolving subtype of stickleback.

From the example of sticklebacks you can see how, when given that the two fish interbreeding results in a half-breed with reduced fitness, you get strong natural selection favoring any genes that cause the two populations of sticklebacks to shy away from interbreeding. Another good example of this may be found on Galapagos island in their finch population. There are several sizes of finch that are each well adapted to feed on a specific type of food when food is scarce. There are finches with large strong beaks that are well suited for the hard nuts that smaller finches cannot crack open. There are finches with long beaks that are well suited for harvesting the seeds from cactus, and there are several others that are equally specialized. As you can imagine, if any two of those finch subtypes bread with each other, you would get offspring that are in between niches and aren’t as well suited to feed on either. Thus this is another example where evolution is being set up to take place. Although all of the finches are still able to breed, natural selection is favoring genes that cause them to not want to breed, because in doing so they reduce the fitness of their offspring. They in fact do exhibit preferential mating, and scientists have noticed that they have evolved different songs. Since birds attract each other for mating with their songs you can see how selection for that trait would end up increasing the birds fitness.

Now that I have described several examples of sympatric speciation known to evolutionary biologists that are currently taking place, I will attempt to concisely list all the steps involved in achieving evolution via sympatric speciation for the purpose of recapping the main points:

Populations need to have the opportunity to split into sub niches. For example there needs to be distinct niches that allow for the separating populations to fill them so that they aren’t forced into competition before they have gotten a chance to actually separate out into species.
Populations need to have some form of fitness advantage that goes along with developing preferential or physically impossible mating. This could be in the form of a less specialized heterozygote.
Populations need to maintain their lack of mating behavior until the process of natural selection has driven them to being fully incapable of cross breeding. At that point you have separate species.

Pre-zygotic reproductive isolation, post-zygotic reproductive isolation, and physical separation all simply set the stage for natural selection. The same process that drives the alleles conferring whatever form of reproductive isolation of two subpopulations to fixation eventually also drives the two populations into separate niches or perhaps direct competition with each other. In the first of the two possible outcomes you will get two similar seeming but distinct species. In the second example you have the potential for one population to wipe the other out, effectively erasing it from ever having existed. It is by this second process eliminating groups and the first pushing populations further and further apart that we do not see a more complete fan of species ranging from the simplest life forms to the most complex. Instead we see species with seemingly profound differences and wonder how they could have ever shared a common ancestor.

Lose Weight and Stay Healthy in as little as 60 seconds a day!

noreply@blogger.com (John St. John) — Tue, 19 Feb 2008 08:28:00 +0000

Thats right, you heard it! My revolutionary work out plan will turn around your health in as little as 60 seconds a day. The plan's sound reasoning is based on the "anything is better than nothing" concept. Chances are you do not work out, and you probably feel slightly guilty about it. You have probably tried more exhaustive work out strategies only to find several months down the road that your much needed day of rest quickly flushed all memories of your commitment to health. That is where my plan comes into play. Just do something, anything at all, for as short a period as you desire, every day. Forming habits are hard so I say just give yourself credit for doing anything you can, every day.

On that note I start day one of my work out plan. I did 15 push-up/jump-squat things, it took me about a minute to do and it felt like something I can manage squeezing into my busy college life. Granted I am feeling tired due to not working out in... well, I am not sure really. Anyhow that is all for now.

More later on trading options on the stock market, or perhaps I will write a little about evolution, or maybe even more on statistical model selection applied to evolutionary trees.

Crazy Judge

noreply@blogger.com (John St. John) — Sun, 02 Dec 2007 02:41:00 +0000

Judge suspended for jailing court
from BBC News | News Front Page | World Edition
A US judge is removed from his post for jailing everyone in court after no-one admitted to owning a ringing phone.

This is hilarious! I really wish I could have been there for that.

The thuggery behind the harmonious facade - International Herald Tribune, Part 2

noreply@blogger.com (John St. John) — Fri, 30 Nov 2007 16:22:00 +0000

The thuggery behind the harmonious facade - International Herald Tribune: "Ma: Enough of this. I've heard from the 'relevant departments' that people like us are not allowed to make big bucks. We're just doing enough to make a living. The police: We haven't bothered your business, have we? Ma: Really? Unless I remember it wrong, you guys once talked to my partner and said, 'if we see him dealing with your company, your business will end.' The police: That's because you did something we didn't want you to do. Over the last few years you haven't made any trouble for us, so we haven't made any trouble for you. Ma: Is that so? You asked me to come here today. Isn't this trouble? The police: How can you say this is trouble? We're friends. Isn't it O.K. to have a cup of tea together? Ma: It's a pity we're not sitting here as friends. Enough beating around the bush, let's talk about why I am wanted here today. The police: O.K., are you or are you not planning to go to Beijing soon? Ma: I am. I'm flying there tomorrow. Any problem? The police: You have to go? Ma then insists that he is only going for business, and the police reply that if that's the case, they won't try to stop him. But they warn him, for good measure..."

This article makes me question the ethics of investing in the Chinese market. Another important question it brings up is whether or not the Chinese government will be able to continue to allow business in China to prosper or whether this type of harassment of the populace will lead to serious economic issues. In any case I am uneasy about my investments in emerging markets at this point...

The thuggery behind the harmonious facade - International Herald Tribune

noreply@blogger.com (John St. John) — Fri, 30 Nov 2007 15:51:00 +0000

The thuggery behind the harmonious facade - International Herald Tribune: "SHANGHAI: Last October, as Ma Shaofang prepared to travel from the Chinese city of Shenzhen to Beijing to attend a writers' conference, he received a menacing call from the police. Why trouble a businessman who wants to attend a conference? The problem was that as a student hunger strike organizer during the Tiananmen protests in 1989, Ma had a 'dossier' that still trails behind him wherever he goes in China. The Chinese calendar is filled with special dates, 'sensitive moments' whose association with events either historical or current put the authorities on alert and the people on guard. October 2007 happened to be the month of the Communist Party's 17th Congress, a once-in-five years affair whose political significance is such that the capital is locked down, potential 'troublemakers' rounded up and even the airwaves scrubbed with extra vigor by censors whose job it is to see that nothing can sully the image of a serene and clear-sighted leadership. So with that backdrop in mind, the police 'invited' Ma for tea. Ma's account of the meeting, which he recently published, and which was subsequently translated by the University of California at Berkeley's China Digital Times, offers a chilling glimpse of a Chinese reality that few foreigners ever see."

Very interesting article... Definitely a must read.

Intense Term

noreply@blogger.com (John St. John) — Thu, 29 Nov 2007 23:54:00 +0000

I know every college student says it, but this term was intense! Most likely I will say it again next term, and the term after that... Anyhow I have been cramming tons of information to attempt and keep up with my new lab while also trying to stay on top of 3 science based classes. Not an easy task but I love it.

On the computer science front I am currently learning my way around python. Python is a high level scripting language with some pretty cool functionality. My only gripe with it so far is that regular expression handling has to be imported. Granted that there is probably a more elegant way to do the regex stuff I wanted to do, I ended up using about 5 lines of code to go through these files the lab uses for running bioinformatics simulations and pull out relevant data for a new project I am working on. I was not pleased with the relatively confusing commands for dealing with regular expressions either. Other than that I love what I have seen so far in Python. It seems much less confusing to me than Perl.

Thanksgiving was fun. I spent it in Seattle with my girlfriend and her sister's family. Seattle is a fun town and I am definitely adding it to my list of habitable locations.

Thats all for now. It's dead week and I need to get stuff done.

learning perl and feeling nerdy

noreply@blogger.com (John St. John) — Tue, 06 Nov 2007 03:56:00 +0000

#!/usr/bin/perl

print(("a","c","e","h","j","k","l","n","o","p","r","s","t","u"," ") [4,13,11,12,14,0,7,8,12,3,2,10,14,9,2,10,6,14,3,0,1,5,2,10]);

#thats how I roll...

Emerging market boom

noreply@blogger.com (John St. John) — Wed, 10 Oct 2007 05:22:00 +0000

Today I am writing about the booming emerging markets. In the past 5 years or so they have been averaging, as a whole, over 20% gains annually. If you were to invest in the Vanguard Emerging Markets Index fund one year ago you would have already earned 40%.

Taking that kind of hot performance into consideration my risky side wants to throw my money at it. Here is where my conservative economist side comes into play. I really do not know a lot about emerging markets. What are the average PE ratios of those markets? That is a very relevant question, it would tell you how much the current prices are due to betting on the future and how much of the price rise reflects actual growth of the economies.

In theory I can understand a certain degree of price inflation on the stocks of emerging markets. They are just now coming into the world economy as actual players and they represent the majority of the world's population. China and India for example are both considered emerging markets. Imagine for a second what it would be like if the majority of the billions of people in China and India became what we in the West would consider "middle class" socioeconomically. As a whole they would have an unheard of spending power and who knows how far that could go. It is an exciting idea. It would be kind of like buying into the stock market back in the early 1900's in the US except... bigger...

Here is my theory. It is based on nothing other than the idea that there must be some kind of gravity involved with the price of those stocks. If they are actually price inflated as I fear, they could suffer the same kind of market crash as we did in the late 20's. If one were to invest in the US stock market at the height of the boom in the 20's they would have had to wait over 10 years to make their money back after the crash. Is there a crash coming in the emerging markets? When will it happen? If I put my money into emerging markets today would I be in the negative for the next 10 years?

Another thought. We know there is an imminent rise of a middle class in the emerging markets. How do we know that those markets are going to be the ones that profit the most from that class. With the expansion of globalization eventually anyone will be able to do business easily and cost effectively anywhere in the world. It is already happening. That said what is stopping todays giants from growing into those emerging markets and being the ones who capitalize. Just because a company is based in the states doesn't mean that it is confined there. Case in point look at the popularity of Japanese automobiles in the US. They are probably making more money from the car market in this country than many of the car companies in this country are. What then makes investing in companies located in emerging markets so special?

I know that in the long term investing broadly across emerging markets is a very good idea. However is now the time to do it? Can it keep going up? Eventually it will and people who invest in it now who have 20 or 30 years to wait will probably be very happy about that decision.

I have things to do and I am not going to bother editing this now. More on this topic as I learn more about it. I will be sure to post something when I find out a little more about PE ratios and such. If they are low I am in!

Investing in Volitile markets, Surf's up!

noreply@blogger.com (John St. John) — Sat, 06 Oct 2007 06:30:00 +0000

I am going to share an idea I came up with a few weeks ago that has been going well. Here is the background: I am a big fan of investing in stocks, bonds and whatever else. I have been doing it since I was about 12. The thing is though that I am a huge wuss when it comes to the market. When stocks are shooting up I get really excited and throw money into the market. Then when the market falls and I start loosing money I typically panic and pull out. By doing that I tend toward selling on the lows and buying on the highs. I guess some investing is better than none at all. Anyhow I invested in some really secure high yield municipal bonds that provide a slow and steady tax-deferred income stream. That was so boring! I couldn't take it. I am young anyhow so I hear taking more risk at this point in my life is the sensible thing to do. Anyhow I needed to get into a nice risky investment that would still allow me to have enough money to put a down payment on a house when that time comes. The option I came up with: REIT's. They make total sense for my situation and here is why. I want risk and some excitement, REIT's provide that. I want something with high historical returns, REIT's have that as well. I want something that will still give me enough money for a down payment when the time comes. REIT's may be the best option for that. My logic is that if I put my money in something that will track the general trend of the housing market, and say the housing market plummets like so many of us are afraid of, I will lose a bunch of money. The plus side though is that I will not need as much money for a down payment if that happens so I am still OK. Say on the other hand that the stock market as a whole plummets but hosing holds firm. My REIT's will hold to the general trend of the housing market and I will not be hurt as badly by the fall of the stock market as a whole. So as I see it, REIT's although risky are the best option for me and my goals. For what it is worth, that is all for now...

Progress in the new lab

noreply@blogger.com (John St. John) — Thu, 20 Sep 2007 14:33:00 +0000

For the past 2 weeks I have been trying to cram in all the information I can on what is called "Maximum Likelihood". This is a mathematical technique for figuring out, probably among other things, a tree that best represents how a given set of DNA, Protein or whatever is related to each other. From what I gather so far the maximum likelihood method does this by plotting out all the possible trees that could fit the data set and then it computes which tree is the most probable given whatever model parameters you feed in. In other words the maximum likelihood is the highest probably of your data given the best model. A simple example often used is tossing a coin. Say you toss one 100 times and you get 47 heads and 53 tails. Now we have to chose a model. The model is our theory about how the coin flipping works. I will propose 2 and we will see which results in the maximum likelihood. The first model is the "fair coin" model, that you get 1/2 heads and 1/2 tails. Next I will propose a weighted model, say 1/4 heads and 3/4 tails. In the "fair coin" model we get a high likelihood because of the fact that 47/53 is so close to our expected value of 50/50. The second model doesn't seem to be a very good picture of what is really going on, 25/75 is pretty far away from our actual result of 47/53. Thus given the two models I proposed the one that leads to the maximum likelihood given our coin flipping data is the "fair coin" model. That process I just went through is called Model Selection and it is more specifically what I am going to be working on. It is not an exact science and what I gather is that I am going to be working on ways to improve the models used for inferring how much evolution takes place between two sets of data sequences.

When I figure this out to a fuller extent I will be sure to write a more accurate entry on the subject. If you are really interested in it though check out the wikipedia page on Maximum likelihood at http://en.wikipedia.org/wiki/Maximum_likelihood also you could check out the wiki on model selection at http://en.wikipedia.org/wiki/Model_selection.

Take care and have a great day!
-हवे अ ग्रेट दय!

Leaf Cutter Colony

noreply@blogger.com (John St. John) — Mon, 27 Aug 2007 23:23:00 +0000

Cool picture of the ant colony courtesy of Jack Niedbala.