How much money should we save? (1 Viewer)

[Evan S]

No, the "balanced portfolio" mantra infers that bonds are a great diversifier in a balanced portfolio as they have historically a low corrolation to equity movements. Bonds themselves are NOT a riskless option, but used in a portfolio with equities, they can enhance a portfolio return given historical market fluctuations.

Adding ANY asset with a low corrolation to the basket of assets already contained in the portfolio will increase return AND reduce standard deviation (risk) at the same time.

Adding ALL risky assets (as well as riskless assets like US treasuries) creates the market portfolio and efficient frontier (efficient frontier is a portfolio that maximizes return for each given level of risk) as stated by the research of William Markowitz.

 

[PhillJones]

GordonL,
I'm not sure I fully understand that. Since share prices move in a random walk, that fluctuation is just as likely to move in the opposite direction. In which case, you'd have fewer shares. Since it's random walk, one scenario is equally as likely as the other. Unless there's more to this. I know that share prices follow a slightly scewed version of Brownian motion, I don't know how that affects the statistics.

AjayM,
I don't know what job you had while at Univeristy but for many people, investing whilst still in college isn't really a practical proposition. Since most people leave college in debt, they'd be borrowing money to invest at a lower interest rate. It makes much more finacial sence to pay down debt than to invest.

Those trends are on too short a time-scale to give me any confidence. If I'm looking at 30 or even 40 years, I want a trend lasting a lot longer than my investment length, not a factor of two, an order of magnitude. Extrapolating over a similar scale-length to your data is mathematically flawed. It's obvious why people do it. The data simply doesn't exist. There is a definate change in behaviour at several points in history causing apparent discontinuities; fitting across those is meaningless, so we fit back to the last discontinuity and cross our fingers that there isn't another one (and that it'll be a bad one) while we still have our money in the market.

Millions of people may well have managed it but that's the last 40 years. Things change, and the trend of the last 70 may not continue for the next 30-40. It seems to me that people draw too many conclusions from insufficient data.

Not that it's definatley all going to end in tears, but there's no way to know whether it will or won't.

At least, that's the way it appears to me. I'm the first to admit that I don't know much about all this but I'd like to know more before I entrust my life savings to this huge machine and I can't seem to find any financial advisors with a sufficient grasp of mathematics to allay my fears.

 

[PhillJones]

Check out this graph



This one suggests very strongly that 1929 wasn't a discontinuity, the extreme swing just makes it look like it.
It clearly shows that the constant percentage growth model is a good fit for at least 100 years, which isn't too bad but it also shows that, if we accept that the market grows exponentially, it's overvalued and we're probably in for a prolongued period of stagnency.

Is that right or is there more information that I don't know about.

Does anybody know where I can get the raw data. I'd like to fit it myself.

 

[AjayM]

Where do you get exponential growth from? I don't see that on the Dow, that is one of the factors that caused the Nasdaq to "crash". 10%/yr is not exponential, that's steady growth.

Andrew

 

[Philip Hamm]

I'm not an economist and I haven't studied economics in detail, but I have looked into the subject at some length.

Classical economics is based on some assumptions that have no basis in reality. Cheif among these is a cornucopian belief in an endless cycle of growth. I don't buy this, I subscribe to a more scientific concept that there are limits to growth. From what I've been reading in the last few years, I personally believe that we are bumping up against those limits right now.

Just because the stock market has advanced like that for 100 years is no guarantee that it will continue to do so forever, or even for long.

That's pretty far outside the box so to speak.

 

[PhillJones]

AjayM,


If something grows by a constant percentage each year, the the change in value per year is proportional to the absolute value itself.

So,

dx/dt = kx
k is the constant of proportionality

We re-arrange to get
1/x * dx = k dt

Let's integrate it
ln x = kt + c
c is the constant of integration

Raise both sides to e
x = exp(kt)*e(c)

e(c) is the value at t=0; we shall call this x_o
x=x_0*exp(kt)

Now we see that k is our growth factor, and x_0 is our starting value.

Looking at the graph I showed above, we have a semi-log plot (the y axis is logorithmic, not linear). If you plot an exponential function on such a graph you get a straight line.

Philip,
You have similar fears to me then. I'm not sure that we're hitting any limits, but the market looks chronically overvalued to me.

 

[PhillJones]

That depends on the debt, the return on investment, the interest rates involved, etc. But if you are paying 3% for your student loan yet are getting 20% back from your investment portfolio you are much better off paying the debt "normally" (rather than paying it off quickly).

20%? really? That seems awfully high.

I propose from looking at that chart I posted, you were extremely lucky, timing wise. Not that I doubt your investment skill but your were investing during a period of growth which, looking at the chart looks to be well and truely over for (if the pattern continues) the next 20 years.

 

[GordonL]

You're still thinking short time frame. Sure, within any short time period it is quite possible to lose money, which you've observed yourself. Take the example I posed and apply it to any 30 year period using the DJIA as your mutual fund. You'll come out ahead everytime. We're talking long term, regular investing and there has never been a slide that lasted more than 3 or 4 years let alone 30 years. Could it happen? Anything is possible and nobody can predict the future. You have to assess the risk for yourself. And it's not just about statistics - it's about the US and world economies, consumer confidence, world events, etc. Ask yourself what could happen that can cause the market to slide or limit growth for the next 5-10-20-30 years and how likely is it to occur?

 

[AjayM]

20% is attainable today, but it takes some work/luck/dillegence. But the point remains the same even if you lower it to say 8%, however the other items that effect a person's life may affect a decision to pay off a debt or not (say monthly cash-flow). But it's still pretty easy here, you're paying off a 3% loan but earning 8% return on your money in your portfolio, if you have extra cash, where should it go?

Even that is over simplified, just because you're earning 20% in the markets doesn't mean you should go out and finance a bunch of toys or otherwise over-extend yourself, etc just because you can go low/no interest financing. Other than cars and homes and VERY select other items I choose to pay cash for everything, now I may still use a 0%-no payments deal to keep money in my own hands longer, but the cash for whatever I'm buying is still in the bank (so to speak).

 

[PhillJones]

Says who?

The numbers show an over-valuation of about 7,000 points of the DJIA, if we assume the constant growth model on which the whole stock market investment thing is based.

If you're getting 20% in this climate, write a book about it, you'd make even more money. Harvard Universities fund manager doesn't manage those kinds of returns and, apparently, he's pretty good.

 

[AjayM]

Google has only tripled in value in less than a year (as an example). Although I didn't really make it clear, when I mentioned making 20% I went beyond funds and other associated "retirement" type "accounts". I don't think there are any funds out there that are performing to those kinds of numbers for any significant amount of time.

And just to keep things light hearted, I still keep some bucks in this one - http://www.vicefund.com - I didn't really take it seriously at first, but it has done fairly well.

 

[PhillJones]

The graphs I posted above which shows all the data and the trendline. As well as what looks like quartile and limit lines.

 

[AjayM]

So you can tell if a companies value is under or over-rated by a trendline from an index? Talk about needing to write a book! Forget all that good stuff like P/E ratios, market caps, etc.

 

[PhillJones]

Hey, I'm the one, who doesn't think that trends are valid predictors of the future. You're the one who said look at the trend, so I did, and that's what it says.

If that trend line isn't valid then no predictions about market growth being 12% over time etc are valid either.

Those measures you mention refer to individual companies, not the market as a whole. I would guess that an argument could be made that the price of a large number of companies, behave, on agregate according to certain rules that are more to do with the underlying mathematics of market behaviour and less to do with individual circumstance, at least to first order. The same thing is observed by physcholgists. An idividual is unpredictable. A crowd however, behaves in the very predictable ways.

This logic is of course more application to things like mutual funds, rather than individual share shopping which is what you're talking about.

 

[AjayM]

Huh? The indices consist of a group of companies, underperforming companies get dropped and performing companies are added in their place.

What does that have to do with valuation of a public company? How can you see an over-valuation based on a trend line?

Andrew

 

[Bob Graz]

You should save 20-25% of your salary. Roughly half should go into a 401k type fund and the other half in more liquid funds. As you build up your liquid (short-term) assets, you can then start taking more risk in investing with that money.

Don't worry about what you are saving for, house, college for kids, etc. Just save as much as you can and it'll be there for whatever you need it for.

 

[Evan S]

If you think this is true, I suggest you read "The Tipping Point" by Malcolm Gladwell.

 

[PhillJones]

I'm sure you're familiar with the anecdote intended to describe how chaos theory works. The one that says that a butterfly flaps it's wings in Australia and by a complex series of cause and events, it rains in central park.

Does that mean that we can predict the weather by seeing if butterflies flap their wings, of course not. What it means is that the conditions for both rain and shine exist and all it takes is some small input into the system to send it one way or another. Furthermore those inputs are being constantly supplied and not only can we not keep track of them, we don't have to because when it's all added up, the quantity and randomness of the inputs mean that the mathematics of how the system works as a whole wins out. In other words, you can describe how the system works on the whole (the macroscopic scale) but the actions of indivdual people/companies/raindrops cannot be predicted without looking at the individual factors affecting each one (the microscopic/mesoscopic scale).

As for social upheaval or market crashes or whatever, when conditions exist that are ripe for something to happen. Then factors that would not have instigated a change previous, suddenly become empowered. There are always things are people wanting affect all manner of changes.

 

[Mort Corey]

Phill...I've not read up on it enough to have even a basic understanding....but I think you're getting into the area of the Elliot Wave Theory.

Mort

 

[PhillJones]

Thanks Mort, I just googled Ellit wave theory. It certainly looks interesting. Maybe if I get the time and inclanation I'll have a go at doing some kind of wave analysis on the dow.

How do you get hold of the raw data for such a thing?