As much as possible, as soon as possible

Let us start with the second point first. Assume Suzie saves $1k per month I starting at age 25, and stops at age 34. Her friend Bobbie starts at age 35 with $1k per month and saves until 64. Bobbie never catches Suzie.  Assuming 7% return, Suzie's savings grow to $1.3 million, and Bobbie's to $1.2 million.  This is because of what Einstein called the most powerful force in the Universe, compound interest. Or to quote Benjamin Franklin: Money makes money and the money money makes makes money. The inescapable conclusion is to start as soon as you can.

As to the second point, how much should you save?

Dr Johnson: "Resolve to be not poor.  Whatever you earn, spend less."

Whatever you earn, spend less.  By the way, there is a corollary in keeping your weight down: whatever you burn, eat less.

Leaving that aside, how much less than you earn should you spend?  An old rule of thumb is to save 10% of your pre-tax earnings.  I would view this as the very minimum.   The more the better is my view, with a personal target of 20%.  The obvious reason is this allows your capital to accumulate twice as fast, and as we have seen with the compounding discussion above, that is tremendously important.

There is another reason however:  if you spend less, you are used to living on a smaller amount.  Here Dr. Johnson can be helpful again:

"Every man is rich or poor according to the proportion between his desires and his enjoyments; any enlargement of wishes is therefore equally destructive to happiness with the diminution of possession, and he that teaches another to long for what he never shall obtain is no less an enemy to his quiet than if he had robbed him of part of his patrimony.

No. 163 (8 October 1751)."

And this: "Frugality may be termed the daughter of Prudence, the sister of Temperance and the parent of Liberty.

No. 57 (2 October 1750)."

See also the Way to Wealth, a compendium of Benjamin Franklin quotations from Poor Richard's Almanac.

If what you have is sufficient for your needs, material wants will not be the cause of your unhappiness. As stated by Sheryl Crow, it’s not having what you want, it’s wanting what you have.  By saving 20%, you get used to living on the 80%.  If you are used to living on 100%, retirement is going to be a drag.

A numerical example to confirm.  Assume your per annum income is $60k.  We already have the math above that saving 20%, $12k per year, $1k per month yields $2.5 million by retirement.  I discuss spending down assets below, but assume a 3% withdrawal rate, and that gives you an income of $75k per annum.  As you are used to living on $48k per year, that is considerable cushion.  What if instead you save 10%, $6k per year, $500 per month.  You will have $1.25 million at retirement, or an income of $37.5k per year.  But you are used to living on $54k per annum - not fun.

As Dickens put it: “Annual income twenty pounds, annual expenditure nineteen six, result happiness. Annual income twenty pounds, annual expenditure twenty pound ought and six, result misery.”

In other words, keep your nut small. Your nut is your average living expenses. 

A couple of side points.  You could calculate the precise figure between 10 and 20% that would yield the same post retirement income as pre retirement.  But stuff happens, and a cushion is your friend.  The example also entirely disregards inflation, but it also ignore a rising income, which should more than offset the silent tax.

How to save is an interesting point.  In the book The Millionaire Next Door, the richest folks relative to their earnings are careful budgeters, who know where every penny goes.  But the next richest are those who pay themselves first.

How do you pay yourself first?  Direct your savings to your company's 401k plan, and after that to your private savings plan, directly, before you send the rest to your spending bank account.  If you wait until the end of the month, you will find you do not have any extra money to save.  A gentleman named Parkinson had a related statement: "work expands so as to fill the time available for its completion."  In terms of money, you will spend whatever you leave available to spend.

Here we turn to Will Rogers:

“The quickest way to double your money is to fold it in half and put it in your back pocket.”

The above is not intended to slight the value of a budget.  Programs like Mint can easily track where you are spending money.  When you know where it goes, you know where to cut.  Housing is an important area to review when budgeting.  Per The Millionaire Next Door, do not get a mortgage that is more than twice your gross annual income.  The book explores the reasons why at length, but the short version is you are able to save, and you get used to living within your means.  A big house leads to fancier furniture, keeping up with the Joneses vacations etc.   

The only problem with budgeting is it is a pain. It is time consuming, and causes stress.  If you simply set aside the money before you ever see it, you eliminate a lot of the anxiety.  Give the money a few years to compound and you can find saving to be a bit addictive, in a good way.

This is a good point to discuss goals vs habits.  As Scott Adams puts it, goals are for losers.  Either you meet the goal, at which point you need a new goal for incentive; or you miss it, and feel like you failed.  If instead you think in terms of habits or systems, you are making a best effort at a repeated pattern, and over time it will take no willpower - it is a habit.  To paraphrase Warren Buffett, choose your habits carefully, for with time they will become unbreakable chains.

How to Invest 

So now that you are saving 20%, where do you put it?

My favorite book on asset allocation is Unconventional Success: A Fundamental Approach to Personal Investment by David F. Swensen.  Mr. Swensen has run the Yale endowment for decades, and has a prior book on how to do the same.  Unfortunately there are investments available to Yale that are not available to the general public, so Unconventional Success lays out the asset classes available to ordinary folks.

Here are his recommended asset classes as I recall them:

35% U.S. Stock Market Index Fund ((8.11% over ten years, 11.53% since 1928)

15% International Stock Market Index Fund (4.79% over 10 years)

20% Real Estate Investment Trust Index Fund (8.54%)

15% US Treasuries (4.72% all bond index fund, 5.28% since 1928 for US Treasuries)

15% TIPS (Treasury Inflation Protected Securities)

You really should read the book, but I will attempt to summarize why the above asset allocation works.

There are five very different asset classes.  It should be noted that the US Market Index Fund is by far and away the best performer over a large period of time.  Full time scales are not available over the same time frame, at least with a quick google search.

So why not put all your money in US Stocks?  Well, how would you react when stocks drop? (2008 -38%, 2002 -21%, 2001 -12%,  2000 -9%, 1990 -3%, 1981 -5%, 1977 -7%, 1974 -26%, 1973 -14%...you get the idea.  Just for fun: 1929 -8%, 1930 -25%, 1931 -43%, 1932 -9%).

What are the chances you will stay in an all stock portfolio with 4 year losses like that?  I am sure the 3 years in 2000-2002 were not much fun either.  As a side note, down years are a good time to treat stocks like grocery staples - if they are on sale, buy more.  Warren Buffett says it better - in my reading list below, I highly recommend reading his annual letters, now available for the last 50 years. 

Fortunately, diversification is the one free lunch in investing.  I do not recall the book that documents it, but a 90/10 stock/bond allocation beats a 100% stock allocation over time with lower variability.  When one of your asset classes goes down, another may go up or at least hold steady.

As an exercise, during the Great Recession, I took a snapshot of performance of the model portfolio, shown below. 

While clearly not a fun time, it would be a lot better to hold a variety of assets at a time of stress than an all stock portfolio.  Rothschild put it well when he said in effect buy stocks to eat well, bonds to sleep well. 

As you can see from the long run returns (since 1928), this class returns the highest yield over the long run.  A side benefit is that stocks even provides some protection against inflation.  Tobin’s Q (fwiw Tobin won the Nobel) explains that if it is cheaper to buy the underlying assets of a business by buying shares in the firm, people will do so until it is not.

Most people primary invest in their home market, but stock markets do not vary at the same pace.  Look at the 5 and 10 year returns above.  Domestic stocks were down, international up.  This allows you to stay the course.

And it allows you to rebalance.  Periodically (quarterly or annually), sell some winners and buy some losers to keep your target allocations in place.  This allows you to sell high and buy low, without trying to time the market.

First, note that the equity in your home, or in your business properties should be counted in this asset class.  Second, it varies in performance from the other assets, while achieving equity like returns.  Finally, it protects against inflation. 

Bonds are a special case.  While the long term returns are in general not as good as stocks, we do not buy bonds for return ON capital, but rather for return OF capital. In stormy weather, this asset class will always be there if you need the money.

Which is why you should not buy anything but US Treasuries.  Corporate or other bonds have some credit risk.  They can go bankrupt.  The US Government can not.  

Regular treasuries go up in value at times when interest rates are falling.  This is because all assets are valued by taking cash flows divided by discount rates (CF/i).  To learn more, google Net Present Value calculations, but essentially a dollar tomorrow is worth than less than a dollar today (in normal inflationary times).  Since the cash flow is fixed for a bond, the CF is divided by a declining i.  In inflationary times, the CF is fixed, the i goes up, and the bond falls in value.  Interest rates fall when the economy tanks, so bonds go up when stocks fall, as amply demonstrated in the chart above. 

TIPS (Treasury Inflation Protected Securities) are actually the best of both worlds.  They pay a fixed coupon plus inflation, so in inflationary times, they hold their value.  In deflationary times, the coupons fixed portion is steady, and so provides protection.  And the return OF capital is guaranteed.

If balancing all these asset classes seems like too much work, target date retirement funds from Vanguard perform essentially the same function in one fund. However, if you have a private business or significant real estate assets, it would be better to build a portfolio with the handful of asset classes shown above so you can more accurately rebalance. 

Why an index funds for your holdings?  

Costs. Here is a recent quote from the Wall Street Journal:

WSJ 010515: Mean­while, Van­guard is un­der­cut­ting many ri­vals on fees. In­vestors pay 18 cents for every hun­dred dol­lars they in­vest with Van­guard, com­pared with $1.24 for the av­er­age actively man­aged mu­tual fund, Morn­ingstar said. The com­pany also is beating its pas­sive ri­vals, which charge an av­er­age of 77 cents for every hundred dollars.

Now think in terms of $1,000,000 - which is by no means a lot of money when you have your assets accumulate over time.  A regular fund will charge you $12,400 to manage your money, Vanguard will charge you $1,800.  That is a savings of over $10,000 per million - EVERY YEAR.  Which will then compound EVERY YEAR.

This is pretty important, so I will emphasize another way.  A basis point is 1/100th of a percent, i.e. 100 basis points equals 1 percent.  So a regular fund charges 124 basis points, and an index fund charges 18, or about 1/7th as much.  And on average a managed fund does not meet the performance of an index fund.

One more way to look at it: in the future stock returns will likely be lower.  Some Nobel prize winners estimate 7%, or 700 basis points.  So a managed fund takes 124/700 or 18% of your returns.  An index fund charges 18/700 or 2.5%. 

But what if the active management outperforms the index fund?  The funny thing is they do not.  If varies year to year, but index funds mathematically have to beat at least at 50% of the funds, and historically outperform 2/3 or more of funds.  Lower cost, better performance.

And the funds that outperform one year suffer from a law of investing: reversion to the mean.  By all means review the literature, but a fund that outperforms will not continue to do so.

By the way, the above really understates the value of index funds, because they buy and hold stocks, versus the churn common at active funds. This tremendously lowers their tax burden. A brief article that explains this is copied below from Motley Fool. 

Final thought: why?  To be independent. 

Tax Efficiency of Index Funds, From The Motley Fool

A Lesson in Tax Efficiency

The poor performance of the average mutual fund relative to the S&P 500 on a pre-tax basis has been highly publicized over the past few years. A story that has been told a lot less is that on an after-tax basis, the returns get even worse. Index funds outperform money managers not only through minimizing expenses, but by staying invested and not changing those investments very often. Simply looking at the capital gains distributions in an index fund relative to other professionally managed money provides an object lesson about how low turnover and low cash balances can provide solid returns.

Because the Standard & Poor's 500 is an index constructed by the editorial board of Standard & Poor's, index fund managers have an advantage most professional money managers do not. Other than the cash they might need for redemptions in the fund, the question of what stocks to buy is just a matter of looking at the weightings in the S&P 500 and distributing the money. The question of what to sell is even more elementary, as the index fund only needs to sell when the index is changed. On average, due to mergers, acquisitions, bankruptcies, and general corporate distress, the S&P 500 can have anywhere from five to fifteen changes during a given year. Compared to the average mutual fund, however, five to fifteen "sells" is an extraordinarily low number.

Looking at the capital distributions in the Vanguard Index Trust 500 for 1997, every share saw 14.5 cents of short-term capital gains and 40.5 cents worth of long-term capital gains. Given that the Vanguard Index Trust 500 closed the year with a net asset value of $90.07 per share, you can see that this meager 55 cents in taxes barely reduced the returns in either fund. Of course, if an investor had sold shares of the fund, they would have to recognize tax consequences beyond the distributions. However, for individual investors creating wealth over long periods of time through the magic of compounding, on an after-tax basis the more money you can keep invested, the more value you can create. Just as every dollar in fees you fork over to invest hurts your returns, every dollar you give to the tax man does just the same.

By way of comparison, Vanguard's Windsor fund saw a short-term gain distribution of 86 cents and a long-term distribution of $2.02. With a net asset value of $16.98 at the end of 1997, Windsor generated 6.4 times the tax burden per share but had a net asset value that was 81% lower. Not only did Windsor only return 21.98% in 1997 versus 33.35% for the S&P 500, but much of what it did return was automatically taxed through capital gains distributions, shaving 15% to 36% of those returns off the top depending on the investor's tax bracket. Although most mutual fund companies say that they do not stress after-tax returns because each individual tax situation is different, if after-tax returns were reported the performance gap between the average mutual fund and the S&P 500 would increase substantially.

An ancillary benefit of not selling very often and having a preset investment plan is that you can keep the amount of money you have in low-yielding cash to a minimum. An index fund can keep 98% to 99% of its money invested, whereas the cash position of the average mutual fund is much higher. With only 90% to 95% of your money invested, you actually need to beat the index return with the money that has been invested to match the return of the index in your fund. So to match the return of the index, money managers not only need to outperform in order to make up for higher expenses, but they also have to outperform because they cannot be 100% invested. On top of that, even if they do breakeven or beat the market on a pre-tax basis, they can destroy this value for individual investors through capital gains distributions that make most of the gains taxable.

The lesson here for individual investors who want to go it on their own is pretty simple, but pretty powerful. Tax efficiency and maximizing money in the market that is compounding value is a key part of generating long-term, excess returns. In fact, when you calculate the effect of taxes, the margin of outperformance that many short-term oriented investors say they enjoy is whittled down. Although the standard defense is that you have to sell sometime, the reality is that if the money compounds over long periods of time and you capture lower tax rates for being a longer-term holder, on an after-tax basis the difference can be quite profound. That does not even include the possibility if you own a dividend yielding stock that you could just capture the rising income generated by the stock to live on and leave the equity untouched -- and untaxed -- to pass on the higher cost basis to your heirs. This is what the S&P 500 Index is essentially doing, and why it is such tough competition for professionally managed money.

By understanding why the S&P 500 index fund can be a superior investment vehicle, you can gain insight on how to invest in order to be a superior investor. By keeping expenses low, minimizing your tax burden, and maximizing the amount of money in the market, the average S&P 500 index fund kicks the pants off professionally managed money. Everything you can do as an investor to replicate this and create powerful returns on an after-tax basis is something that will create excess returns for you -- returns above and beyond what the market average. In the end, it is these after-tax, excess returns that are the object of the investing. Anything that does not produce similar results has to be questioned.