Heads up, I’m going to talk about my finances a bit here, in case that is some kind of taboo with you.
After being paid a meager stiped for a year, I’ve actually netted a bit of money. When I tutor kids, I charge exorbitant rates1. I think I just got myself onto IKEA’s paid survey short-list. That and I just had my last grandparent die, my grief offset by $104. So I feel like I’m in pretty good shape.
Before, I just had a non-interest checking account which was totally idiot proof. The only things that cost money were other banks’ ATMs and bounced checks. My bank then bought every other bank on the planet so they are all my bank’s ATMs and I definitely don’t bounce checks. But now I have too much money for that account.
First things first, it is time to set up an IRA, Roth, obviously. Jon recommended Vangaurd, which looks like an excellent choice. It turns out that we picked the same fund. So that gets $4000, the maximum.
But now what? Savings account? CD? Interest checking? Anyone else think they’ve got this nailed? Here is Jon’s solution and it sounds pretty good.
But I still need a checking account. I could imagine a future where that is no longer necessary, but that isn’t today. After shopping around, I see this as the winner in the checking department. In fact, it is so good, it becomes hard to justify having a savings account. The extra interest on a savings account wouldn’t cover the time I would spend managing it. Well not until I have a lot more money.
But there are still many thousands of dollars that I don’t plan on touching for at least a year. Potentially longer. So here are the financial concerns, the probability of needing a new car cumulative density function is roughly 1-e^-t, where t is measured in years. There is also the potential for a “starter house” in the 5-10 outlook. In the 20 year outlook there is a dream house but my current investments will be utterly irrelevant if I’m building a million dollar house. For the starter house, the same is probably true. What I do with these $6k won’t really matter, or will they? I think $24k would matter. I would need 15% interest to quadruple in 10 years2.
I am pretty confident in the index funds on which my IRA is based. I am planning on having it beat my mortage. To first order it is roughly like the sampling distribution of the means of N(2,5) [notation: normal(mean,stdev)] quarterly with 4*t datapoints. I know you can’t take the mean of interest rates but it is a decent approximation. I also know that N(2,5) is probably a horrible model, (it is colored and definitely not normal) I have some probability code lying around that I could maybe try. What sort of model do you guys use? I also know that most people probably couldn’t write this sort of code. What do they do?
But then there are mutual funds, some of these look more like N(3,8) or N(4,10). At this point, the normal model probably has to go. But still, I think it is reasonable to expect the variance to go like t-.5. But this is not a stationary process. Yikes.
I also need a distribution of utility of investment growth given time for the next 10 years. Currently, Utility(-6k—6k) has only slightly negative curvature because I am comfortable on my stipend and I still have my parents holding the safety net. The only negative comes from the fact that my parents would have gotten a CD and if I lost money, it would be hard to convince them that I hadn’t acted rashly.
There is some concern that the curvature will get more dramatic in the future if my expenses increase and my parents become less willing to help. But to offset that, once I finish grad school, I expect to make much more money than I do now, which will surely flatten the curve. But on the other hand, I will have more invested. It seems that in absolute terms, I am pretty safe, but I shouldn’t plan on maintaining so high a fraction of my net worth in high risk investments.
So here are some candidates.
0.1) checking account 3% interest
0.2) CD 4-4.5% interest
1) I could put more money into the IRA fund. For now at least, the taxes I would pay are very little and certainly not enough to bring it down to CD level.
2) Index fund,
- whole market ,
- S&P
3) Regional mutual funds
- Domestic,
- Europe,
- EAFE,
- World 1,
- World 2,
- Emerging
4) Sector mutual funds.
- Financial,
- Energy ,
- Real Estate
- Biotech? Health care?
What do you guys think? Any recommendations? Any better ideas?
1 The Academic Skills Center’s tutoring rate is absurd. They want me to tutor a kid a $7/hour? What are lectures $100/hour? I can usually promise a equal or higher dot{grade} (and sometimes also dot{knowledge}) than the lectures. How much does the person who sends out the ”’tutors badly needed for physics 13’ emails” make?
2 Here is how you do a log calculation in your head. If ever you are hanging out with management and you pull this out of your ass, they’ll be impressed. I have heard of this move winning a job interview.
- you have to be able to do log base two in your head. If you have a choice (like I do here) just pick a convenient one.
- memorize the natural log of 2 = .693
- taylor series of ex = 1 + x + ...
- convert your log base 2 to base e by means of (2)
- divide by your time constant.
- compute (3) (i.e. add 1)
But for interest rates, the 1 is implicit anyway so you can skip 6)
example, what interest rate doubles my money in 30 years?
log2(2) = 1, ln(x) = .693*log2(x), ln(2) = .693
ln(2)/30 = .02
answer: 1.02 which is typically called 2% interest.
Strickly speaking 14% is too far to use the first order taylor series of ex with only the linear term. To be more precise, compute the next term in this case, .142/2 = .01 so the answer is more like 15%.