follow me on Twitter

    Blog has moved to, see August, 2007 on new site

    Blog entries for August, 2007

    This post moved to

    I don't think John Markoff at the New York Times knows about the GNOME Online Desktop effort, but Owen noticed this blog post where Markoff works around a hard drive crash with a Linux LiveCD and web-based apps, just as we've proposed. He calls it Computing in the Cloud.

    What I discovered was that - with the caveat of a necessary network connection - life is just fine without a disk. Between the Firefox Web browser, Google’s Gmail and and the search engine company’s Docs Web-based word processor, it was possible to carry on quite nicely without local data during my trip.

    I had already stashed my almost 4,000 sources and phone numbers on a handy web site which I had access to, and so I found the only things I was missing were the passwords to online databases and my files of past reporting notes and articles which I occasionally refer to.

    Bouncing between hotel rooms to Wi-Fi-enabled lobbies and conference rooms, I was easily able to stay online and file my stories without incident.

    Afterwards it made me wonder why there aren't more wireless, Web-connected ultralight portables for business travelers. Somebody, it would appear, is missing an obvious market opportunity.

    - John Markoff

    I posted to desktop-devel-list about the current password-storage problems in GNOME.

    This post moved to

    While I'm on financial topics, if you don't know how to do "time value of money" math you should stop and learn about it right now. For all I know I'm the only person who didn't learn this in high school, but in case there's someone else, here's a blog post.

    Because of inflation and ability to earn interest, a dollar today is not the same as a dollar next year. To compare two financial choices (say, renting a house vs. buying a house), you need to convert the choices to money at the same point in time, just as comparing two measurements requires them to be in the same units.

    The buttons that make a calculator a financial calculator are the ones that let you convert any series of cash flows (money spent or earned) to the value of that series at a point in time.

    The calculator buttons are:

    • n - number of periods of time
    • i - interest rate per period of time
    • PMT - a regular per-period cash flow
    • PV - present value, i.e. value of all PMT and FV at time 0 assuming rate i
    • FV - future value, i.e. value of PV and all PMT at time n assuming rate i

    A financial calculator can solve for any of these given the other four.

    In a spreadsheet, there's a separate function depending on what you want to solve for. From the OpenOffice docs:

    • PV(Rate; NPER; PMT; FV; Type)
    • FV(Rate; NPER; PMT; PV; Type)
    • NPER(Rate;PMT;PV;FV;Type)
    • PMT(Rate; NPER; PV; FV; Type)

    The "type" is 0 if PMT comes at the end of the periods, and 1 if it comes at the beginning.

    Next time you have a financial decision, try to break it down into cash flows and you may find you can get a handle on it with these formulas.


    • The calculations are sensitive to the interest rate you pick. If you start picking overoptimistic rates you will make all kinds of bad decisions.
    • In addition to only comparing money amounts at the same point in time, you can only compare money amounts if they are both in real dollars or both in nominal dollars. That is, be sure to reduce your interest rates to consider inflation when appropriate.
    • These formulas assume a steady rate, which is very misleading for stocks and other risky assets. Actual results will be in a wide range. When planning retirement savings, you need to run the numbers with a very pessimistic rate of return, not only with an average one. You can also find software that uses monte carlo simulation to show a spread of possible returns.
    This post moved to

    This week's Fortune magazine has a scary black cover and says "Market Shock 2007." The market drop was 10%! Not very large in historical terms, and it comes after a huge gain. But from the news (in Fortune and almost everywhere else) it sounds like we should be freaking out.

    I thought I'd post the point of view I find helpful when making personal investment decisions and trying to keep perspective. If you are averse to reading about financial stuff, stop reading. Like most of my points of view, this one is really lengthy.

    Disclaimer: this is not investment advice for your personal situation, and if you plan your finances around what you read on a software developer's random Internet blog post you deserve what you get. Do your own research. You have been warned.

    That said, here are some thoughts, and then some elaborations.

    I'll elaborate on some of these points.

    First, investing probably isn't the most important financial concern for you. Attempts to get abnormally high investment performance are even less important.

    Lots of media articles are about beating the market. But beating the market is a dumb goal most of the time.

    Unless you have an awful lot of money to work with, beating the market on a consistent basis by 1% (very hard) or 5% (virtually impossible) will still add up to much less than, say, a 5% raise in salary. If you focused on a career change or a promotion, you could probably get 20% more. Do the math. You're better off worrying about your career - or starting a business - if your goal is to become dramatically richer.

    Aside from the income you earn through work, some other important non-investing finance topics to worry about include: 1) having proper insurance (long-term disability, life, health, liability/umbrella, etc.) 2) running a budget surplus 3) having an emergency fund 4) taxes 5) having a will and power of attorney. All those areas probably matter more than an extra couple percent over the market. Here is a sensible book on all that. Or ask for help from someone who charges an hourly fee instead of commissions.

    Second, unlike beating the market, investment risk does matter, and you need to worry about investing at least enough to fully understand the risks you are taking on and why.

    If you need the money in a year and stocks drop 20%, you are screwed, even if you "beat the market" by 5% since the market dropped 25%. Or if you really can't deal emotionally with a 50% drop (I know I can't), but you take that risk anyway, then even with a longer time horizon you'll experience plenty of angst that keeps you from focusing on more worthwhile activities.

    Here's what you should spend time on: have a solid investment policy that you know is right for you and that you understand. Know when you need the money, and choose appropriate investments for the time horizon.

    Choosing a stock or fund that beats the market by 2% won't do a lot to change your life. Choosing an appropriate investment policy, however, makes a very big difference.

    Once you have a budget surplus so you're saving enough, the most important investment policy decision by far is simply how much in stocks, how much in bonds or cash. 80/20, 60/40, etc. You should spend as much time on this decision as you need to be 100% comfortable with it, then forget about your investments until you need the money.

    Benjamin Graham has a quote I keep at the top of my investment policy:

    An investment operation is one which, upon thorough analysis, promises safety of principal and an adequate return.

    He defines the terms like so:

    • thorough analysis: "the study of the facts in light of established standards of safety and value"
    • safety of principal: "protection against loss under all normal or reasonably likely conditions or variations"
    • adequate return: "any rate or amount of return, however low, which the investor is willing to accept, provided he acts with reasonable intelligence"

    My personal risk tolerance is low. I am comfortable taking moderate, informed risks after reading a lot and feeling I have a very good understanding of what I'm getting into.

    But this is a personal thing. The luckiest people are comfortable taking sensible risks without having to spend a lot of time reading, and so they can get on with their life. Other people are most unhappy if they miss out on big gains, and don't mind losing money as long as everyone else is losing money too. And of course circumstances vary - maybe you need college tuition next year, maybe you don't need the money until you retire in 40 years.

    You have to make a personal, informed decision. Don't let anyone else do it for you or tell you what you should think.

    But do try to vividly imagine what it would feel like to experience the investment losses you sign up for. If you see on that your plan could lose 40% in a year, then do the math: multiply 40% by your investment balance and write that number down and imagine losing it overnight.

    Third, the stock market is not just a random bouncing number; over the long term it does relate to "fundamentals" (real economic forces).

    (Here we get into my personal need to study the risk in detail before investing. If you have no such need you may want to skip the rest of this post.)

    The premise of every (sane) investing strategy is that over the long term - and that means at least 10 and more like 20-30 years - on average stock market performance will reflect economic growth, and historically economic growth has been positive, in countries with no major crises. The market reflects economic growth because individual stocks reflect the value of businesses, and businesses have a value because they make money and their owners get the money they make.

    Without this premise, index investing won't work, value investing won't work, basically nothing would work reliably and investing in stocks would amount to gambling.

    With this premise, there's a long-term upward trend line (fundamental value), and the market's current price bounces around it.

    You probably know about the "random walk" concept, that the return on the market tomorrow is random and unpredictable, given conditions today. This is absolutely true over short timeframes, like a day or even a year. But as far as I know, the theory is not intended to imply that investment returns are truly random over the long term. Investing would make no sense if returns were random over 30 years.

    Reasonable investment strategies assume that there is a trend upward, on average. This upward trend happens because stocks represent something of actual value, the value increases over time due to economic growth, and the value is roughly measurable.

    (This is not a given. Commodities, for example, do not trend upward on average, after inflation - they are just random bouncing numbers. Unlike stocks, there's no fundamental reason pork bellies or metals would be worth more in ten years than they are today, other than inflation. Oddly, commodities can still be worth investing in, but only in small doses combined with something that goes up, like stocks.)

    Fourth, if we are willing to assume there's a long-term upward trend, there are a number of investment strategies that should reflect this trend and produce gains on average over time.

    What these strategies have in common is that they work for a reason, due to economic growth, businesses producing cash profits, and just plain math. They aren't "magic."

    The simplest strategy is to buy a passive index fund and wait. Over 30 years, the bouncing around the trend line doesn't matter.

    It can matter a lot over shorter timeframes, though. Say you bought at the 2000 peak, you made zero dollars since then, 7 years later. And 7 years later things are still overpriced. (That's how insane the tech bubble was. Ouch.)

    Some minor elaboration helps a lot. The two elaborations most people use are dollar cost averaging and asset class diversification (combined with periodic rebalancing). Both of these reduce vulnerability to market fluctuations.

    A more complex elaboration that sells more when the market is above the trend and buys more when it's below is value averaging.

    Sensible active management strategies work too.

    Graham/Buffett style value investing follows a discipline that hopes to earn something close to the long-term upward trend, on average, over time. But it can lead to very different behavior from the market as a whole in the short term, for example many managers using this discipline missed out on most tech bubble gains and then also missed out on most tech bubble losses.

    John Hussman's fund, which I mentioned earlier, uses analysis based on historical statistics to vary the amount of market risk taken from 0% to 150%, where at 0% money market returns are expected. This fund would have been painful to own the last three years, since it's earned about 5%. But then again, it destroyed the index in 2001 and 2002. The fund's record is too short to feel completely confident it will have long-term returns in the same range as an index fund, but the strategy does make sense and "should work" based on history, in the same sense that index and value investing "should work."

    It's controversial whether these strategies are worth it vs. an index fund, since it isn't clear these strategies can beat the market.

    My opinion: the value of an actively-managed fund comes primarily from risk control, and from the psychological knowledge that if there's another tech bubble (or mortgage bubble), there will be some stuff in my portfolio that isn't tangled up in it. I would generally expect the actively-managed funds I like to lag a bull market and do better in a bear market, but the purpose of investing in them (for me) is not to beat the market overall.

    Vanguard suggests 50% in actively-managed funds and 50% in index funds and that's what I use.

    Tangents aside, the point is: sensible strategies are based on the premise that there is a "fundamental value" - an economic reason that investments will grow. This fundamental value can be captured "actively" or "passively," but any strategy without a basis in fundamentals is junk. Or at least too risky for me.

    Fifth, sensible investment strategies avoid predicting the future.

    Strategies that work reliably make the minimum prediction: that fundamental value will grow over the decades, roughly resembling historical experience. Then they have a discipline which still works even if nothing else is predictable. They can handle "statistically impossible" events like the tech bubble or the 1987 crash and come out OK, because they did not rely on predicting those events.

    Nassim Nicholas Taleb has a very interesting book about how unpredictable the world is. He isn't even willing to rely on fundamental economic value growing over time, preferring to invest mostly in cash. I don't see the need to go that far - to hedge the apocalypse, buy food and guns, forget about stocks. (The range of future situations where stocks are permanently worthless but T-bills remain valuable seems small to me.)

    Nonetheless, I like Taleb's argument that nobody can predict any complex event. And successful investors like Buffett make a point of not trying to, using primarily past events - proven profits and current prices - as input to their decisions.

    Hussman emphasizes this point over and over as well, for example:

    Frankly, I don't know whether investors will drive the market even higher in the weeks ahead. My opinion is that whatever gains emerge (and indeed, much of what has already emerged) will ultimately prove quite temporary. What I do know is that certain factors have reliably identified egregiously bad times to accept market risk, and that every historical instance similar to the present has been a disaster. The current instance may very well prove to be the exception, but I do not invest shareholder assets on the hope that the future will be entirely at odds with all available historical evidence.

    Passive indexing, value averaging, and rebalancing are all based on this same premise: they are systematic approaches that on average are helpful, according to historical experience. In any given instance, if you could predict future price movements, you could do better than these strategies. But you can't predict.

    This is why it's solid advice to "buy and hold" and "don't time the market."

    I'm very skeptical of mutual funds that are based on predictions, such as macro economic analysis or guessing at which companies will have the hot new technology. Nobody can consistently get those predictions right, or at least I can't tell which fund managers have this superpower.

    Sixth: If you believe my first five points, and chose your investment policy well to reflect your goals and risks, you have no reason to worry about anything you read in the media.

    To prove the point, let's split the media into two categories.

    Category One is made up of people who agree in broad outline with a couple of my points here: that investing is about capturing a long-term increase in fundamental economic value, and that nobody can consistently predict the future.

    Category Two is made up of short-term traders, numerologists, economists, and the like - all trying to predict the future, or even worse, the short-term future of market prices.

    If you read Category One pundits, many are arguing that the market is overpriced. I quoted some of them at the start of this post.

    However, you, like them, should already have a disciplined plan that produces adequate return, with acceptable risk, even if the market is overpriced sometimes and underpriced sometimes. You don't need to know the market price next week or next year. Because nobody can predict the future, there is nothing to do except stick to the disciplined plan.

    Sure, if there's a true apocalypse or a government coup, you are screwed. But like I said, if that keeps you up at night, buy some guns and food to hedge your stocks.

    You can ignore Category One commentators, or treat them as interesting but not worrying, because you're already prepared. Your investment strategy may already include adjustments to changing market prices, through averaging or hedging or allocation; or it may not; but in any case you've already planned for market prices to fluctuate and know what action to take (if any), or know what action your fund managers will take. The actions you've planned will give you the right risk and return over time.

    You can ignore Category Two punditry because it's meaningless garbage. Nobody in Category Two ever relates what they say to a sensible investment strategy, or spells out the long-term statistics that back up their points. They're a lot like news anchors who report on politics "horse race" style, focused on polls, and never analyze the substantive policy issues the elections are supposed to be about.

    A month ago, there were two popular media claims about why the market would keep going up. The first was "liquidity" and the other was "the forward P/E ratio is in line with historical averages."

    Impressive-sounding comments, but meaningless for investing.

    "Liquidity" is a short-term (and hand-wavy) property of the market. The historical average return for stocks covers many periods where liquidity came and went. If you say stocks can be priced more highly when liquidity is favorable, you're assuming that liquidity will stay favorable permanently (or that you can "get out" in time if it doesn't and "get back in" at the right time after you get out). The last month is a good demonstration that liquidity won't stay favorable forever.

    The price/earnings argument is bogus in two ways. First, it uses the so-called "Fed Model" to argue that because interest rates are low now, stocks are worth more. (Hussman argues that historical statistics don't support this.) Like the liquidity argument, this assumes that either rates will stay low or that you can get out and back in if they rise, neither of which has to be true.

    Second, the price/earnings argument ignores the business cycle. Profit margins go up and down over time. Like liquidity and interest rates, earnings can and probably will become much worse at certain times and you will not be able to market time when it happens. If (for example) all this housing turmoil creates a recession, earnings will go down, and if the price/earnings ratio stays constant, prices will go down too.

    Right now, profit margins are at a record high. The price/earnings ratio is average, but if you change the profit margins to also be average, the price/earnings ratio is very high.

    So in early July, liquidity was good, earnings were at record highs, and the 10-year Treasury had a low-ish rate. Several favorable signs.

    Investing based on this information is like saying "I don't need to own an umbrella because it's not raining right this minute." When people who believe in long-term fundamental value say stocks are overpriced, what they mean is that among other risks, liquidity could dry up, profit margins will go down eventually, and interest rates could rise. And the risk of those things happening means that on average, stocks are likely to go down again. Good conditions don't continue unbroken for 30 years.

    I like to read Category One media sources to keep myself from believing in Category Two media sources. Category One is good at emphasizing that Category Two can't predict the future and has no sensible long-term strategy. Here are some sources I like:

    Sensible, fundamentals-based commentary keeps the focus on risk, not return.

    Again, never understand anyone's opinion as a prediction. It is never appropriate to try to "get in" or "get out" of the stock market, based on today's commentary or an emotional feeling. Only systematic disciplines have a good chance of working.

    There is no way anyone will read this post to this point, but if you did, I hope you found some part of it helpful.

    [1] A caveat about Morningstar: they are best known for the star ratings for funds, which in my opinion are useless and often a recipe for performance-chasing. However, their "analyst picks" and qualitative fund writeups are very good, and they also do fundamental analysis of many individual stocks, allowing them to analyze the whole market by combining all the bottom-up analyses together. You have to buy Morningstar Premium to get at some of the more useful content.