Something happened at the New York Stock Exchange last Monday. It was a market tantrum borne out of months of speculation about economic instability in the world, potential interest rate policy changes at the Federal Reserve, seemingly tepid economic growth in the USA and who knows what else. I have been watching the value of my retirement portfolio decline over the last 9 months or so, from its peak in December of 2014. It has declined 10%, to be exact. Against that, I have collected dividends of approximately 2%, all of which have been reinvested, suggesting my share count has risen by 2% while the value of those shares has dropped by 10%.
The question is, should I care? No one likes to see their net worth shrink. On the other hand, every company I own continues to pay dividends, generally rising dividends. Cheaper shares mean I purchase a bit more every quarter when dividend-reinvesting time comes around. My brokerage statements suggest my dividend receipts are rising, although I don't keep a separate spreadsheet. One of the hard things about watching that value drop is that I have also invested another 25k or so into the total portfolio at the same time. Is this one of those "times that tries men's souls"? Not quite, I would venture, but it's certainly a time when I am questioning whether I should be DRIPing or collecting cash. Had I collected that 19k in cash dividends from the first of the year, I could deploy it now, at 10% lower cost than I did over the last 6 months. Had I saved that 25k in new contributions, I could also deploy that at lower cost than I did over the last 6 months.
If you only buy when there is "blood on the streets", then there is a lot of time when you aren't buying. That would mean very large cash positions building over time. One would have to have a strategy for cash. In my IRAs, I could be using cash to sell puts and produce more cash. However, the brokerage fees cut at least 10% out of short term put premiums.
The 19k dividends and 25k new contributions add up to 44k investment. Had I invested at 10% lower cost, I'd own $4400 more in stock, or about 0.4% of the portfolio, or probably not enough over which to commit ritual suicide. In this next 6 months, I'll acquire some stock at lower prices and perhaps invest new contributions at lower prices as well, depending on where things move from here.
As it was, I was on my way to work when the flash-crash was occurring and by the time my workday was done, it was over, halfway recovered, and I didn't have cash to deploy anyway. For me, it was an event that wasn't. I sat on my hands, so to speak. There wasn't anything to do, other than to either pay attention or not.
It was a busy week, so I didn't. By week's end, things have further recovered. We'll see what the next week brings. I'm DRIPing at lower prices and no new contributions are flowing into the portfolio at the moment. Another boring week at the retirement portfolio....
Saturday, August 29, 2015
Monday, August 17, 2015
Another run at the concept
I am in the accumulation phase of my retirement investment portfolio. I would like to retire with adequate dividends and distributions from my retirement accounts to pay expenses. The date I can cover those expenses is my Paycheck Emancipation Date. It may or may not be my actual retirement date, but it represents the earliest time I can voluntarily choose to retire. One way to get a rough fix on that point in time is to know the Dividend Doubling Time of my portfolio.
Various metrics inform the investor of the performance of one's portfolio. You already know most of them, so it doesn't pay to recite them all. Investors in the dividend growth space have eschewed following the value of the portfolio from day to day, as well as the total return. Some of us continue to sneak a peak at those numbers, although our better selves know that anything connected to the price of a stock is subject to the vagaries of the market, rather than strictly related to the business performance of the underlying companies.
By appropriate selection criteria, one can assemble a basket of equities representing companies who have paid dividends faithfully over years, and an even more select collection of companies who have raised dividends reliably over many years. Dividends are unique amongst ALL metrics in that they are ALWAYS positive, even when other performance metrics may vary from year to year. Even a dividend cut still yields a positive dividend, unless that cut is of the most-feared dividend elimination variety. For most investors, a dividend freeze puts the company on probation and a dividend cut results in a prompt sale and re-deployment of assets.
Since most companies that pay dividends tend to maintain and periodically raise them, a portfolio with multiple such companies will regularly experience increases in the composite dividend income.
Each company within the portfolio will contribute variably to that dividend growth according to multiple factors, but the trend should generally be up.
In the accumulation phase of a retirement portfolio, dividends are generally reinvested, either via DRIP, selective reinvestment or both. The compounding effect of dividend reinvestment accelerates dividend growth within the portfolio.
Finally, most investors who are contributing to a retirement portfolio contribute regularly to the account, so additional infusions of cash are invested, producing additional dividends as well.
So, the three components to dividend growth are dividend raises by individual companies, the compounding effect of dividend reinvestment and addition of shares via cash contributions to the account over time.
There will be no single metric that accurately estimates the rate of dividend growth within the portfolio, so one must measure it in real dollars on an ongoing basis, or estimate it by back-of-the napkin methods. Still, it's a very important metric to the dividend investor, as it represents the real cash flow produced by the investments themselves, once cash contributions are converted to equities. All of us hope to produce a sufficient stream of cash that we won't be forced to sell shares to pay the bills.
So, how does that back-of-the napkin estimation of dividend growth work? First, new contributions amount to a percentage in the growth of the share count. I contribute an additional 5% to the corpus of my retirement account every year, so if it is broadly distributed across the equities I own, I know that next year's dividends will grow by roughly 5%, other factors notwithstanding. Second, the composite current yield of my portfolio is between 3-4% and I reinvest all dividends via my brokerage's DRIP program, so I know that next year's dividends will increase by slightly more than that composite yield, due to the compounding effect of either quarterly or monthly dividend reinvestment. That puts me between 8-9% dividend growth per year within the portfolio. To be sure, this is not the pure investment performance of the portfolio, because the majority of the growth is due to new purchases. However, in my real world, it still represents a growth in the cash-producing engine that will be my primary source of income in retirement, so every contribution matters.
Finally, there is a composite rate at which the companies I own increase their dividend payments. This varies quite a bit from company to company, so I don't have a great fix on the actual number, but it is safe to estimate that it adds something short of an additional 1% to the actual growth in cash flow from dividends. After all, if I take my 8-9% more shares every year and add intrinsic dividend growth, I'm pretty sure that adds up to less than 10% intrinsic dividend growth per year in the aggregate.
This rough estimation turns out to be pretty close to the truth, as I look at the brokerage statements from my retirement account. The actual dividends in the whole portfolio are growing at roughly 9-10% and my dividend doubling time looks like it is a bit under 7 years. If I choose Uncle Sam's definition of retirement age for SSI, I have about 12 years to go earning paycheck, so I'm expecting just under two doublings of my dividend income before that date. There are a bunch of reasons I expect I won't continue to draw a wage that long, so I'm taking that into account as I plan my investment contributions.
Getting a rough fix on the nominal dividend growth rate within the portfolio allows one to predict the rate at which the dividend will double, or the Dividend Doubling Time (DDT). For any given equity within the portfolio, the dividend doubling time will vary by the rate at which the company's executive decides to adjust the dividend year by year, as well as decisions to purchase new blocks of stock. This makes calculation of an individual equity DDT fairly useless, except to say that it will be less than the time estimated by simply predicting a doubling of yield on cost of the initial position.
I get lost in the weeds when another writer does a comparison of two companies with different yields and dividend growth rates, projecting the point of dividend parity many years into the future and discussing the merits of the one stock over the other. I am very skeptical that this type of analysis is meaningful, understanding that any number of macroscopic economic factors may intervene and multiple events could occur within a single company that might change it's dividend growth over many years time. I'm much more comfortable with estimating what may occur to my entire portfolio over the next 3-5 years if I continue on the same investment pathway, understanding that I am monitoring and shaping the portfolio with my purchases.
Have you calculated that ACTUAL rate at which the dividend stream in your portfolio is growing? Do you have a rough estimate of the dividend doubling time, or DDT of your retirement portfolio? How does that inform your decisions about the next purchase?
Various metrics inform the investor of the performance of one's portfolio. You already know most of them, so it doesn't pay to recite them all. Investors in the dividend growth space have eschewed following the value of the portfolio from day to day, as well as the total return. Some of us continue to sneak a peak at those numbers, although our better selves know that anything connected to the price of a stock is subject to the vagaries of the market, rather than strictly related to the business performance of the underlying companies.
By appropriate selection criteria, one can assemble a basket of equities representing companies who have paid dividends faithfully over years, and an even more select collection of companies who have raised dividends reliably over many years. Dividends are unique amongst ALL metrics in that they are ALWAYS positive, even when other performance metrics may vary from year to year. Even a dividend cut still yields a positive dividend, unless that cut is of the most-feared dividend elimination variety. For most investors, a dividend freeze puts the company on probation and a dividend cut results in a prompt sale and re-deployment of assets.
Since most companies that pay dividends tend to maintain and periodically raise them, a portfolio with multiple such companies will regularly experience increases in the composite dividend income.
Each company within the portfolio will contribute variably to that dividend growth according to multiple factors, but the trend should generally be up.
In the accumulation phase of a retirement portfolio, dividends are generally reinvested, either via DRIP, selective reinvestment or both. The compounding effect of dividend reinvestment accelerates dividend growth within the portfolio.
Finally, most investors who are contributing to a retirement portfolio contribute regularly to the account, so additional infusions of cash are invested, producing additional dividends as well.
So, the three components to dividend growth are dividend raises by individual companies, the compounding effect of dividend reinvestment and addition of shares via cash contributions to the account over time.
There will be no single metric that accurately estimates the rate of dividend growth within the portfolio, so one must measure it in real dollars on an ongoing basis, or estimate it by back-of-the napkin methods. Still, it's a very important metric to the dividend investor, as it represents the real cash flow produced by the investments themselves, once cash contributions are converted to equities. All of us hope to produce a sufficient stream of cash that we won't be forced to sell shares to pay the bills.
So, how does that back-of-the napkin estimation of dividend growth work? First, new contributions amount to a percentage in the growth of the share count. I contribute an additional 5% to the corpus of my retirement account every year, so if it is broadly distributed across the equities I own, I know that next year's dividends will grow by roughly 5%, other factors notwithstanding. Second, the composite current yield of my portfolio is between 3-4% and I reinvest all dividends via my brokerage's DRIP program, so I know that next year's dividends will increase by slightly more than that composite yield, due to the compounding effect of either quarterly or monthly dividend reinvestment. That puts me between 8-9% dividend growth per year within the portfolio. To be sure, this is not the pure investment performance of the portfolio, because the majority of the growth is due to new purchases. However, in my real world, it still represents a growth in the cash-producing engine that will be my primary source of income in retirement, so every contribution matters.
Finally, there is a composite rate at which the companies I own increase their dividend payments. This varies quite a bit from company to company, so I don't have a great fix on the actual number, but it is safe to estimate that it adds something short of an additional 1% to the actual growth in cash flow from dividends. After all, if I take my 8-9% more shares every year and add intrinsic dividend growth, I'm pretty sure that adds up to less than 10% intrinsic dividend growth per year in the aggregate.
This rough estimation turns out to be pretty close to the truth, as I look at the brokerage statements from my retirement account. The actual dividends in the whole portfolio are growing at roughly 9-10% and my dividend doubling time looks like it is a bit under 7 years. If I choose Uncle Sam's definition of retirement age for SSI, I have about 12 years to go earning paycheck, so I'm expecting just under two doublings of my dividend income before that date. There are a bunch of reasons I expect I won't continue to draw a wage that long, so I'm taking that into account as I plan my investment contributions.
Getting a rough fix on the nominal dividend growth rate within the portfolio allows one to predict the rate at which the dividend will double, or the Dividend Doubling Time (DDT). For any given equity within the portfolio, the dividend doubling time will vary by the rate at which the company's executive decides to adjust the dividend year by year, as well as decisions to purchase new blocks of stock. This makes calculation of an individual equity DDT fairly useless, except to say that it will be less than the time estimated by simply predicting a doubling of yield on cost of the initial position.
I get lost in the weeds when another writer does a comparison of two companies with different yields and dividend growth rates, projecting the point of dividend parity many years into the future and discussing the merits of the one stock over the other. I am very skeptical that this type of analysis is meaningful, understanding that any number of macroscopic economic factors may intervene and multiple events could occur within a single company that might change it's dividend growth over many years time. I'm much more comfortable with estimating what may occur to my entire portfolio over the next 3-5 years if I continue on the same investment pathway, understanding that I am monitoring and shaping the portfolio with my purchases.
Have you calculated that ACTUAL rate at which the dividend stream in your portfolio is growing? Do you have a rough estimate of the dividend doubling time, or DDT of your retirement portfolio? How does that inform your decisions about the next purchase?
Sunday, August 9, 2015
Is DDT a DUD?
So, off I went on the little voyage of discovery, calculating dividend doubling times for a few of my holdings.
Now, it's not simply a doubling of the dollar value of a year's dividends that interests me. You can look that up in a decent brokerage webpage or elsewhere. I'm interested in the time it takes for my actual dividend payments per year to double, taking into account the incremental increase in dividends paid as well as the effects of dividend reinvestment on the cash flow within the position year by year.
As usual, I did this in a "quick and dirty" fashion by looking at actual dividends paid, quarter by quarter, for companies I have held over several years. It turns out, I have made incremental investments in most of them, so I do not have a position that has been intact long enough to double purely on the basis of dividend increases and dividend reinvestment. One can calculate a rate of dividend growth from one year to the next, however, and it's interesting to note that most of my holdings have dividend growth rates of between 7-12%. This is achieved by dividing the dividend received in a given quarter from one year into the dividend received in the prior year. A number a bit greater than one appears, from which you subtract the number 1 and multiply the result by 100 and you have the rate of growth in percent. Using the rule of 72, you can divide 72 by the number you get and the doubling time will appear. My dividend doubling times range from 5-10 years for most stocks, less for a couple. The dividend doubling time, averaged across the portfolio would be a great number to have, as it predicts the time at which I might possibly receive adequate dividends to replace the income I earn with my labor.
Just as the dividend growth rate varies from year to year, the dividend doubling time varies as well.
It's important to realize however, that the published rate of dividend growth under-estimates the dividend doubling time if you are a DRIPer, because your stock count is growing at the rate of the current yield in addition to dividend increases programmed by company management.
A high dividend company like T will experience a dividend doubling of 14.4 years on a static dividend. You don't have to raise the dividend by much to reduce the doubling time to under 10 years, due to the dominant effect of a dividend reinvestment of 5% every year.
Finally, you can't ignore the "new money effect" of additional contributions to the portfolio. That isn't investment performance in it's purest sense, but since I am the investor, the rate at which I save is a piece of my portfolio growth performance. I have been adding to my retirement portfolio at the highest rate allowed by the law and will continue to do so until I stop earning a paycheck. That currently is adding about 5% per year to my nest-egg. If you treat the new money like additional dividends, adding the current portfolio dividend yield to the new addition, gives a dividend rate of approximately 8%, which in itself produces a dividend doubling time of 9 years.
When I look at the actual dollar figures, new money plus dividends is adding 9.5% cash to the portfolio every year, and the dividend is rising by 10% per year. That results in a dividend doubling time of 7 years.
I don't think DDT is a DUD. It tells me about where I'll be in 7 years with respect to cash flow derived from my retirement portfolio.
Now, it's not simply a doubling of the dollar value of a year's dividends that interests me. You can look that up in a decent brokerage webpage or elsewhere. I'm interested in the time it takes for my actual dividend payments per year to double, taking into account the incremental increase in dividends paid as well as the effects of dividend reinvestment on the cash flow within the position year by year.
As usual, I did this in a "quick and dirty" fashion by looking at actual dividends paid, quarter by quarter, for companies I have held over several years. It turns out, I have made incremental investments in most of them, so I do not have a position that has been intact long enough to double purely on the basis of dividend increases and dividend reinvestment. One can calculate a rate of dividend growth from one year to the next, however, and it's interesting to note that most of my holdings have dividend growth rates of between 7-12%. This is achieved by dividing the dividend received in a given quarter from one year into the dividend received in the prior year. A number a bit greater than one appears, from which you subtract the number 1 and multiply the result by 100 and you have the rate of growth in percent. Using the rule of 72, you can divide 72 by the number you get and the doubling time will appear. My dividend doubling times range from 5-10 years for most stocks, less for a couple. The dividend doubling time, averaged across the portfolio would be a great number to have, as it predicts the time at which I might possibly receive adequate dividends to replace the income I earn with my labor.
Just as the dividend growth rate varies from year to year, the dividend doubling time varies as well.
It's important to realize however, that the published rate of dividend growth under-estimates the dividend doubling time if you are a DRIPer, because your stock count is growing at the rate of the current yield in addition to dividend increases programmed by company management.
A high dividend company like T will experience a dividend doubling of 14.4 years on a static dividend. You don't have to raise the dividend by much to reduce the doubling time to under 10 years, due to the dominant effect of a dividend reinvestment of 5% every year.
Finally, you can't ignore the "new money effect" of additional contributions to the portfolio. That isn't investment performance in it's purest sense, but since I am the investor, the rate at which I save is a piece of my portfolio growth performance. I have been adding to my retirement portfolio at the highest rate allowed by the law and will continue to do so until I stop earning a paycheck. That currently is adding about 5% per year to my nest-egg. If you treat the new money like additional dividends, adding the current portfolio dividend yield to the new addition, gives a dividend rate of approximately 8%, which in itself produces a dividend doubling time of 9 years.
When I look at the actual dollar figures, new money plus dividends is adding 9.5% cash to the portfolio every year, and the dividend is rising by 10% per year. That results in a dividend doubling time of 7 years.
I don't think DDT is a DUD. It tells me about where I'll be in 7 years with respect to cash flow derived from my retirement portfolio.
Sunday, August 2, 2015
Eureka, I've found my own personal investment performance benchmark!
Tonight, I was reading one of my favorite authors on Seeking Alpha commenting on another author's article about the relative contribution of dividends to investment performance over time. He made the point that if you reinvest dividends, over time an increasing amount of total return is attributed to the dividend-derived portion of your position. That assumes, of course, that you purchase an original position and then tabulate the portion of your total return that comes purely from the capital gains attached to your original purchase, and attribute all additional returns (capital gains and dividends) from stocks purchased with dividends to the dividend-derived return.
Interestingly, since the goal of the dividend-growth investor is a rising stream of dividends, an interesting benchmark is the dividend-doubling time, or DDT as I have named it. After all, if one wants to live on dividends and distributions alone in retirement, the faster one gets to replacing one's earned wages with dividends and distributions, the sooner one can ditch one's work clothes for a bathrobe early every morning.
The rule of 72 defines how quickly a pot of money doubles if it earns compounded interest.
7% per year takes 10 years to double. 10% per year takes 7 years to double.
What rule describes the doubling time of dividends? First, one has to set out some requirements for the rule. One rule applies if one takes the dividends as cash. Another rule applies of one is reinvesting the dividends. Applied to a single stock, one must use a DRIP to determine the DDT for that stock.
Within a portfolio, the DDT could be achieved DRIPing every stock, or collecting dividends and reinvesting selectively.
For the sake of simplicity, I will assume the rule applies to a single stock in a DRIP program.
Simply stated, whatever the cash yield of a holding at the time you establish your position, the point at which you are collecting 2x that amount of cash, you have your Dividend doubling time. That time is closely related to a concept called yield-on-cost. If your starting yield is 3% and your yield on cost is 6%, then your dividend stream should have exactly doubled in that interval.
We all know that yields don't double as a company adjusts it's dividends over time. When earnings rise, stock price generally rises as well. When a discrete dividend increase occurs, it has an effect on stock price as well. Since using a DRIP requires that you purchase incremental shares at whatever price prevails at the time the dividend is paid, the incremental purchases have differing claims on future earnings and dividends. If one presumes a constant yield, i.e. the company raises dividends exactly in proportion to rising earnings, and the market re-prices the stock precisely in proportion to rising earnings, then one can relatively easily calculate the point at which the actual cash dividend payment doubles. However, we have this phenomenon of margin expansion and contraction that undulates somewhat unpredictably over time, so a formula will only approximate the real-life DDT of a given stock at any time in history.
Still; the concept is important because real money is what we use to pay our bills; Currently, the cash yield of all my stock holdings is roughly enough to pay the mortgage on my home. If I want to retire on dividends and distributions alone, I need to reach a point where those dividends cover all my expenses, after taxes. I'll need about 4-fold more dividends, or two doubling times. How long will that be? The simplistic view is to assume that my current yield will stay about the same, so I'll need to quadruple the value of my portfolio in order to retire. I think that won't actually be the case. I think that the combination of dividend growth, as well as the compounding effect of dividend re-investment will get me there faster than a simple compounding calculator would predict. However, since rising stock price will dampen the effect of the DRIP in regards to the rate of increase in share count one might expect from the DRIP, the calculation must take that into account. That means that things like share-repurchases, which drive earnings per share and share price may also have to be considered in the equation. Rising earnings per share drives rising dividends per share, as long as the payout ratio remains constant. It may not be constant, however. That is a decision made in the board room, not by some formula. Formulas will never do more than approximate reality, so watching the actual cash numbers makes more practical sense than relying on formulas.
Now I'm off to figure out the dividend doubling time of a few of my holdings! I'll be reporting back soon....
Interestingly, since the goal of the dividend-growth investor is a rising stream of dividends, an interesting benchmark is the dividend-doubling time, or DDT as I have named it. After all, if one wants to live on dividends and distributions alone in retirement, the faster one gets to replacing one's earned wages with dividends and distributions, the sooner one can ditch one's work clothes for a bathrobe early every morning.
The rule of 72 defines how quickly a pot of money doubles if it earns compounded interest.
7% per year takes 10 years to double. 10% per year takes 7 years to double.
What rule describes the doubling time of dividends? First, one has to set out some requirements for the rule. One rule applies if one takes the dividends as cash. Another rule applies of one is reinvesting the dividends. Applied to a single stock, one must use a DRIP to determine the DDT for that stock.
Within a portfolio, the DDT could be achieved DRIPing every stock, or collecting dividends and reinvesting selectively.
For the sake of simplicity, I will assume the rule applies to a single stock in a DRIP program.
Simply stated, whatever the cash yield of a holding at the time you establish your position, the point at which you are collecting 2x that amount of cash, you have your Dividend doubling time. That time is closely related to a concept called yield-on-cost. If your starting yield is 3% and your yield on cost is 6%, then your dividend stream should have exactly doubled in that interval.
We all know that yields don't double as a company adjusts it's dividends over time. When earnings rise, stock price generally rises as well. When a discrete dividend increase occurs, it has an effect on stock price as well. Since using a DRIP requires that you purchase incremental shares at whatever price prevails at the time the dividend is paid, the incremental purchases have differing claims on future earnings and dividends. If one presumes a constant yield, i.e. the company raises dividends exactly in proportion to rising earnings, and the market re-prices the stock precisely in proportion to rising earnings, then one can relatively easily calculate the point at which the actual cash dividend payment doubles. However, we have this phenomenon of margin expansion and contraction that undulates somewhat unpredictably over time, so a formula will only approximate the real-life DDT of a given stock at any time in history.
Still; the concept is important because real money is what we use to pay our bills; Currently, the cash yield of all my stock holdings is roughly enough to pay the mortgage on my home. If I want to retire on dividends and distributions alone, I need to reach a point where those dividends cover all my expenses, after taxes. I'll need about 4-fold more dividends, or two doubling times. How long will that be? The simplistic view is to assume that my current yield will stay about the same, so I'll need to quadruple the value of my portfolio in order to retire. I think that won't actually be the case. I think that the combination of dividend growth, as well as the compounding effect of dividend re-investment will get me there faster than a simple compounding calculator would predict. However, since rising stock price will dampen the effect of the DRIP in regards to the rate of increase in share count one might expect from the DRIP, the calculation must take that into account. That means that things like share-repurchases, which drive earnings per share and share price may also have to be considered in the equation. Rising earnings per share drives rising dividends per share, as long as the payout ratio remains constant. It may not be constant, however. That is a decision made in the board room, not by some formula. Formulas will never do more than approximate reality, so watching the actual cash numbers makes more practical sense than relying on formulas.
Now I'm off to figure out the dividend doubling time of a few of my holdings! I'll be reporting back soon....
Subscribe to:
Posts (Atom)