Archive for the 'FAQ' Category

What are good money market funds?

February 1, 2010

Question: I agree completely w/ moving funds into cash, as your latest newsletter recommends.  I noticed the hyperlink for cash, and following it, see that “cash” means a money market fund w/ a 2% assumed return.  Where are you banking?  The money market funds are all returning the federal funds rate right now, about 0.1%.

I also see in your list that your don’t have a tbills fund in there.  I’d recommend running ishares’ SHV [iShares Barclays Short Treasury Bond Fund] through your algorithm, possibly as an alternative to cash.

Response:

MMF, not “cash”. I should clarify it in the newsletters that “cash” means “money market fund”. Thanks for pointing that out.

Assuming a 2% return. As you know, in November, I modified the ETF newsletter to make its performance verifiable by a company called TimerTrac. For the newsletter to be verifiable, the recommendations I give have to be consistent with the choices that they allow. They give me four choices as to what “money market fund” means. Of the four, I think only two make sense for the newsletter: cash with no interest or MMF with a 2% assumed return. I am going with the latter.

Interest rates vary depending on economic conditions. Sometimes, like now, they are well below 2%. Other times, they are well above 2%. For example, consider the historical performance of Vanguard Prime Money Market Fund (VMMXX). Over the past 5 years, its annualized return has been over 3%; since its inception in 1975, its annualized return is over 6%. Because of this, I think that over the long-term, 2% return is a more accurate representation of money market rates than 0%, and that’s what I’m going with.

As for current rates, according to Bankrate, you can get a savings account with as high as 1.70% APY. Though, of course, it’s tricky to send your money back and forth between a bank and a broker.

SHV. That I only have two bond ETF’s in my list is also due to the limited choices offered by TimerTrac. In addition, it makes things easier to have one fund in the list that is completely risk free. SHV is not risk free as it can and does decline in value.

  • Share/Bookmark
FAQ. Tags: , , , , , , , , .. 0 Comments

Personal Investment Performance (PIP) calculation

January 7, 2010

It’s that time of the year again — people are going to be receiving statements from their retirement plans (including the Thrift Savings Plan) showing how much money they have made (or lost) within the past year.

The confusing part is that, even if you had stayed within a particular fund for the entire year, the return shown on the statement will likely be different than the return of the fund. This happens because people add money to their retirement accounts throughout the year. Every time you add money, you buy a little bit more of the fund at the then current price. This is known as dollar cost averaging or DCA.

What’s more, even if two people were in the exact same fund for the whole year, the returns shown on their statements will likely be different from each other. This is because the return depends on several factors, such as:

  • the amount of money in the account at the beginning of the year;
  • the amount of regular contributions; and
  • the timing of the regular contributions.

Since these three factors are likely to be different for different people, the returns they see on their statements will also be different. Because of this, the return is often called the personal rate of return (PRR) or the personal investment performance (PIP).

Below, I discuss two formulas for calculating this personal return. The first formula, which I call the personal rate of return, is relatively easy to use and to interpret, though it is an approximation. The second formula is the one used by the TSP and other retirement plans. The formula is commonly called the modified Dietz method, while the TSP simply calls it the personal investment performance. Though the formula is exact, I think it is more difficult to use and interpret.

Personal rate of return. The personal rate of return (PRR) is calculated as the weighted average of two other returns, the lump sum total return (LSTR) and the dollar cost averaging total return (DCATR). These are explained below.

Let

  • MB be the money in the account at the beginning of the year; and
  • let MC be the total amount of money contributed throughout the year (that is, it’s the amount contributed per pay period multiplied by the number of pay periods).

Then,

PRR = (MB x LSTR + MC x DCATR) / (MB + MC) – 1

For example, suppose that,

  • at the beginning of the year, you had $50,000 in your retirement account;
  • you contributed $500 per pay periods for 26 pay periods (1 pay period every two weeks);
  • the lump sum total return (discussed below) was 1.2; and
  • the dollar cost averaging total return (discussed below) was 1.1.

Then, the calculation is as follows:

  • MB = $50,000
  • MC = $500 x 26 = $13,000
  • PRR = ($50,000 x 1.2 + $13,000 x 1.1) / ($50,000 + $13,000) – 1 = 17.94%

Now, let’s discuss LSTR and DCATR.

Lump sum total return. “Total return” means 1 plus the return. For example, a return of 20% corresponds to a total return of 1.20. “Lump sum” means the total return on the money that you had in your account at the beginning of the year. For a fund, LSTR is equal to the price on the last day of the year divided by the price on the last day of the previous year.

Dollar cost averaging total return. This is the total return on the money contributed throughout the year. The formula is similar to that for LSTR. It is the price on the last day of the year divided by the effective purchase price. The effective purchase price is the harmonic mean of all the fund prices throughout the year. For details, see a more thorough explanation with graphs and examples.

Personal investment performance. We now come to the formula used by the TSP, called either the modified Dietz method or the personal investment performance (PIP).

Let

  • EMV be the market value of your account at the end of the year;
  • BMV be the market value at the beginning of the year (this is the same as MB above);
  • CF be the total cash flow or contribution into the account (this is the same as MC above);
  • i be the index on specific contributions to the account;
  • CFi be the amount of the i-th contribution (in other words, CF = sum CFi); and
  • Wi be the number of calendar days from the i-th contribution until the end of the year divided by the number of calendar days in the year.

Then,

PIP = (EMV – BMV – CF) / (BMV + sum (Wi x CFi) )

… And that’s how the personal returns are calculated and why they are different for each person. :)

  • Share/Bookmark
FAQ. Tags: , , , , , , , , , , , , .. 2 Comments

Locking in losses

January 6, 2010

I regularly receive emails from people who tell me that they have been following some investment methodology, such as buy and hold, have recently lost a lot of money, have found our newsletters, want to follow our newsletter, but are afraid. (Boy, was that a mouthful. :) )

They are afraid that by switching out of their losing investments, they will “lock in” their losses. They feel that, while they have lost money, those are “paper losses”. But once they switch out of the losing funds and start following our newsletters, they will somehow make those losses more real.

I think that many people get confused for the following reason. They see a fund that has lost a large amount of money fairly quickly, so they reason that it should regain that money just as quickly. They are afraid to get out of the fund because they are afraid to miss this big up move.

The reality is that a big loss in the recent past does not imply a big gain in the near future. The false belief that this implication exists is an example of a cognitive bias called the gambler’s fallacy. Someone who commits gambler’s fallacy thinks that returns are negatively autocorrelated, that is, that negative returns are typically followed by positive returns. However, returns are not autocorrelated, either negatively or positively.

This incorrect belief is why people are afraid of “locking in” their losses. They are afraid to miss the big up move which, they believe, is imminent. However, there is simply no evidence that returns behave in this way.

  • Share/Bookmark
FAQ. Tags: , , , , , .. 0 Comments

The bubble in bonds (or, only the present exists)

January 6, 2010

Question: Although hundreds of billions are going into bond funds at this time, that seems imprudent to me.  Why?  Historically low Fed Rates are bound to be raised.  Even anticipation of rising rates results in falling bond prices; actual rate rises will probably cause even higher rate of bond price drops. In light of the impending bond bubble burst, is it prudent to hold the high percentage of TSP assets in F Fund (“Total Bond”) that your January newsletter recommends for Conservative Allocation?  I’d rather be early in eliminating or significantly reducing my TSP portfolio’s risk that investing in bonds entails.

Response: This is a great question. You and I seem to have different approaches to investing. You are attempting to predict or anticipate what will happen. I am simply trading based on what is actually happening right now. You are saying that rates are bound to be raised and so bond prices are bound to fall. I am saying that bond prices are going up right now, and so that’s why I am in bonds.

The issue with my approach is that things can change — they do not have to stay the way they are. My defense against this is that I update my calculations twice a month, and in doing so, hope to catch things before they change too much.

The issue with your approach is that your predictions might not come true. Since your trades are based on your predictions, if your predictions do not come true, that calls the trades themselves into question.

  • Share/Bookmark
FAQ. Tags: .. 0 Comments

Scheduling of the TSP newsletters

December 18, 2009

Questions: When are the TSP newsletters emailed?

Response: The first newsletter of the month comes out on the Sunday before the first Monday of the month. The second newsletter of the month comes out two weeks after that.

If you are expecting a newsletter to come out around the 1st and the 15th of the month, then, in some months, you might think that they are late. But they’re not.

For example, let’s take December 2009. The first Monday of the month was December 7, which means that the first newsletter of the month came out on December 6. The second newsletter is scheduled for December 20, which is two weeks after December 6.

  • Share/Bookmark
FAQ. Tags: , , , , , .. 0 Comments

“Black swans” and trading

December 18, 2009

Question: What are your thoughts on Taleb and how his work relates to your trading method?

Response: Based on what little I know about Taleb, his main criticism of statistical modeling is that it does not take into account rare but devastating events (“black swan” events). My response is that this is an excellent criticism of how statistics is often done, but not of statistics itself. Rare but devastating events can and should be accounted for in statistical models. Good statisticians do this. But most people who practice statistics are not good statisticians.

As this relates to my investment method, I do think that my statistical model accounts for some “black swan” events. Not all of them, of course. Part of the nature of a “black swan” event is that it is difficult or impossible to foresee.

Not foreseeing things is bad. That’s true. But with a statistical model that you’ve put a lot of thought and work into, you can foresee more than without it. You can always make improvements and do better. But you cannot do the impossible, such as anticipating events that are impossible to anticipate. And that’s OK.

  • Share/Bookmark
FAQ. Tags: , , , , .. 0 Comments

Too much trading?

November 18, 2009

Question: Alex, I have a question.  I see that last week you recommended XOP and XLP for the conservative portfolio.  These ETFs still perform outstandingly, however, today you recommend switching to EWM.  I noticed that XOP went up 2.11%, XLP went up 0.56% today whereas EWM went up 1.55%.  VEA went up 1.56% (VEA was the sole recommended ETF last week for the balanced portfolio).

Are you just switching randomly through different growth sectors even though the currently invested ETF is still going strong?  This would increase commissions and costs.

Response: It’s a great question.

When everything is going up, a strategy that switches often and a strategy that stays in its investments for a longer time will perform about the same. The difference will be, as you correctly point out, that the strategy that switches often will incur more transaction costs.

The problem is that we do not know ahead of time what the future will bring. We do not know ahead of time whether, in the future, everything will be going up. So your criticism, while understandable, applies only when looking into the past, not into the future.

Since I do not know what the future holds, my strategy attempts to invest in what it calculates to be the “best” investment at the time. Even if “the market” as a whole starts declining, hopefully, the “best” investment will not decline, or will not decline by as much.

By the way, the recommendation from last week (Conservative: XOP/XLP; Balanced: VEA) was actually in force for two weeks straight.

  • Share/Bookmark
FAQ. Tags: , , , .. 0 Comments

Our investment method in more details

October 21, 2009

Question: I have been looking to get better returns from my TSP.  However, I have trouble following any method without knowing how or why it should work (much like you state in your bio).  I could not find any solid reasoning on your site. [Could you please describe your method.]

Response: There is a general framework for making investment decisions called “asset allocation”. This framework is popular with academics because, unlike technical analysis indicators, for example, it has a theoretical basis. However, the framework itself leaves out some details that you need in order to actually implement it. My method is an implementation of asset allocation, in which (I think that) I have worked out these details.

Here is how asset allocation works. Each investment has two characteristics, one called expected return, and the other called risk. Intuitively, here is how they are defined. If you took a large number of unrelated investments with the same expected return, then, on average, they would give you this expected return. Risk is a less clear concept. Usually, risk is taken to mean the standard deviation of logarithmic return, which is also called “volatility”. This is a good definition of risk because it is easy to understand — higher uncertainty in future returns means higher risk. However, the definition has some weaknesses. I have developed another definition of risk which I use in my algorithm.

Both expected return and risk are estimated from historical price data. In other words, mine is a “technical” system — it only uses historical price data. It does not use any “fundamental” information, such as earnings. The argument for technical analysis as opposed to fundamental is well known. Even if you think that fundamental information is relevant, by the time you, the investor, hear of it, a lot of other people have already heard of it and acted on it. Thus, any relevant fundamental information is reflected in the price before you even hear about it. By analyzing price, you can know all the relevant information. Plus, of course, technical systems are easier to test on historical data. Simulated performance is one of several pieces of evidence that your system works. See below for more on this.

(Just to clarify, even though asset allocation is a technical approach, it differs from most other technical analysis systems. Most technical analysis uses indicators, such as moving averages, or chart patterns, or other similar methods. However, there is no theoretical basis for those approaches.)

Since expected return and risk change with time, how to properly estimate them is a very big question. Many asset allocation methods, such as the one used by the Lifecycle funds within the TSP, actually assume that these investment characteristics do not change with time. I think that this is a major mistake.

Just as each investment has an expected return and risk, so does every combination of investments. Asset allocation says to pick the mix of investments that maximizes the expected return while not exceeding a fixed maximum risk. Of course, there is a huge number of investment combinations. Thus, it is not possible to check all of them using a “brute force” approach. However, with a good search algorithm, you can find a mix of investments with a high expected return that does not exceed your fixed maximum risk. That’s all there is to it. :) In my newsletters I give two allocations. The difference between them is that they have a different fixed maximum risk.

  • Share/Bookmark
FAQ. Tags: , , , , , , .. 0 Comments

Lifecycle allocations, strategy, and performance

September 21, 2009

Question: How can I find the distribution of the 2010. 2020, etc plans? Could investing partially in these plans result in a similar investment strategy and outcome?

Response: The distributions of the Lifecycle funds change gradually over time. It’s fairly straightforward to figure out what they are at any given point in time. But you have to remember that whatever distributions you find, they will change with time.

The investment strategy of the Lifecycle funds differs from my strategy. Even though the distributions of the Lifecycle funds change with time, these changes are pre-calculated a long time in advance. We don’t know exactly how long in advance, but I suspect it’s years.

This means that the Lifecycle funds do not respond to changing market conditions. When stocks become more risky, the Lifecycle funds do not respond by shifting money out of stocks. When stocks become less risky, they don’t respond by putting more money in.

The Lifecycle funds accomplish what I call “static diversification”. They are diversified in the sense that they are invested in several different markets. But this diversification is static because it does not respond to changes in those markets.

By contrast, my approach does respond to changing market conditions. When stocks are more risky, I put no or little money in them. When their risk declines, I put more money in them.

In my opinion, the performance of the Lifecycle funds is relatively poor. The funds with the longer time horizon, such as L2040, have most of the money in stocks. Thus, they have had performance similar to the performance of stocks, including the huge recent drop. The funds with the shorter time horizon, such as LIncome, have most of the money in the money market. While this is a lot safer, it also limits the potential return.

For example, since the beginning of 2007, assuming you have not made any contributions to your account, the returns are as follows:

  • LIncome: +7.08%
  • L2040: -10.90%
  • C Fund: -19.66%
  • G Fund: +10.98%
  • Share/Bookmark
FAQ. Tags: , .. 0 Comments

ETF and TSP letters contradictory

September 21, 2009

Question: In the two new newsletters, I note that the analysis of the direction of stocks is contradictory.  In the ETF letter, stocks are said to be the thing for the next week while in the TSP newsletter, stocks are seen as risky over the next two weeks.  Given that the only difference in the two letters is generally the number of changes that can be accomplished economically, would it not makes sense to wait one week before instituting the changes in the TSP allocation or am I missing something?

Response: I also saw that my algorithm was giving what looked like strange advice this week, so I investigated further. The reason for the differing recommendations is the time scale. Because the TSP newsletter is updated twice a month, while the ETF newsletter once a week, the TSP calculations use a longer time scale.

On the longer time scale used by the TSP newsletter, stocks look overbought. However, on the shorter time scale used by the ETF newsletter, stocks look close to being overbought, but not there yet.

About waiting a week with the TSP newsletter. I recalculate the optimal allocations for the TSP on the first and third Monday of the month. I have not found any benefit to attempting to time the market. That is why I simply recalculate the optimal allocations on a regular basis. There is nothing magical about first and third Monday — it’s just convenient and it splits the month into two approximately equal parts.

  • Share/Bookmark
FAQ. Tags: , .. 0 Comments