Hi Animorph,
Would you mind elaborating on this strategy? Is it a fixed dollar amount that needs to be exceeded? I am not at the withdrawal phase yet and just assumed I would have to increase my bonds in preparation for retirement, but it seems you decided to go with stocks/cash? Just hoping that you wouldn't mind sharing how you came to the no bonds decision.
Thanks.
This will be long, but I've never really described the whole thing before.
During 2007 - 2009 my main concern was putting more money into cheap equities, to the extent that I used home equity line of credit money to add to my equities at what turned out to be the bottom of the market. Some people on the forum here didn't even want to rebalance excess bond balances into equites to maintain their normal allocations during the same time period. So I don't think my strategy would work for too many here.
I dislike holding bonds and rebalancing to a fixed target because bonds (normally) don't return as much as stocks. For a 30 year retirement, that's a bunch of money sitting in bonds for 30 years. I don't want to do that. Same thing for cash of course. So I'm nominally 100% equities, though very diversified there.
I expect the value of my portfolio to increase in value over time, faster than inflation, even with my withdrawals. That means I'm pulling out a smaller percentage of my portfolio when I'm 100 than when I started, even assuming no other income. What could be safer than that? Why should I be all in bonds at age 100 if I only need to pull out 4% per year? When we die and the estate goes to the kids, does it make sense that that we were heavily in bonds for 10 to 30 years and then the kids inherit, who should move it back equities then?
But I still need to account for the swings of the market. I retired (DW is still working but could have retired with me) in 2007, with severence pay into 2008. Wanting to avoid the retirement horror story of the bear market at the start of retirement, and already concerned about the obvious housing problem and the less obvious consumer spending funded by a negative savings rate, I pulled out 3 years worth of cash to start. That was actually way more than 12% of the portfolio since I was paying for out of state college for one DS (and now the next DS) and of course covered that as well. No SS or pensions yet, but there will be. The idea was to spend the cash until it was gone and then to sell equities. That does not impact lifetime returns too much and provides significant early retirement safety (thankfully it worked). I'm about 7% off my portfolio peak value (including cash), with some moderate withdrawals.
From there, I projected portfolio values at the end of each year using simple average market gains as paart of my retirement plan. In those calcs I sell equities at the start of the year to cover expenses for the year, just to be conservative. What I want to do now is try to even out the market swings that are not part of the plan (though FIREcalc says I'm at a reasonable SWR also). So if my portfolio value ever exceeds the coming end of the year projected value, I sell equities to match the projected year-end value and essentially remove that cash from the portfolio. This might happen often during the year. It was a daily thing in the most recent upturn, if the market went up I sold since I was already at the desired end of year portfolio value. That keeps my plan on track, actually ahead of plan since the portfolio has hit the target value and I have extra cash that was not planned. That cash will be used for expenses, delaying the need to sell equities. So if the market is above "average" I'm pulling out cash.
If I run out of cash (some of 2009 and 2010 as a matter of fact) I sell equities as needed (any time of the year), and may try to avoid unnecessary expenses if the portfolio is particularly low. I tried to leave as much as I could in equities as the market bottomed and started to rise.
If I have medium or long-term cash available (> 1 year expenses) and we hit a bear market, I'll put that back into the market in chunks that are sized to run cash out (reserving the 1 year expenses) at my best estimate of the market bottom. So I was investing all the way down in 2008-2009 from -20% to -40% with existing cash and below that with HELOC funds, all roughly equal chunks at equal percentage steps. We delayed expenses, used up the reserved cash for living expenses while the market finally improved, and then finally had to begin selling equities. By that time stocks were well above the bottom and we had some nice gains already from the cash we had been investing as it went down. It never felt bad to sell at that point, even though we were still way down from the peak.
That just leaves a bear market with no excess cash to spend to contend with, one that does not happen after an over-exuberant market bull where I've pulled out lots of cash. In that case I'll just be taking a hit. If it occurs late in retirement I expect I'll already have increased spending (no bears), have more flexibility to cut spending, and have fewer years of retirement to fund (smaller portfolio is OK). Early on I'll still have some flexibility in spending during a long retirement and some hope of a quick market recovery instead of a permanent hit. That's the risk I take for running closer to the optimum median outcome instead of the optimum safest outcome.
Right now I've got about 1 year of cash pulled out, some of which I used to pay off the HELOC since that offered a better return than just holding cash. I expect to withdraw that money from the HELOC before selling equities again sometime in 2012 if things stay as is. If the market goes up more than about 2% or so I'll be selling equities again and raising more cash. I'll probably stop selling if my cash balance exceeds about 3 years, but keep topping-off to keep it at 3 years if the market is still over target.
So I'm trying to even out market swings while staying nominally 100% in equities. It's easy to tell when you should buy equities (in a bear market). I am using my average market gain projections to know when maybe the market is a little overvalued and I should sell. Barring those conditions, I'm just selling equities as needed to meet expenses.
I'm in cash because it is easy for short-term moves (no trading commissions or short-term redemption fees) and I don't expect to hold significant amounts for long periods. You could also use bond funds or ETF's. The key is that you're always spending it down to zero before selling equities. There is no permanent allocation to cash or bonds. Sort of a double bucket approach with a zero permanent allocation to the cash bucket. Probably sensitive to your average market gain assumption too.
That's it. It's strictly my own design, so there is no article that describes it and gives backtested or Monte Carlo results. I'll find out how it works in about 30 years. Subject to changes before then if I find something better.