I am frequently asked by financial advisors whether clients should "hurry up"—in light of the declining richness of guaranteed living benefit (GLB) riders—and add money to variable annuities (VAs) they already own, before it is too late and they are closed to new contributions.
To be honest, I too, have been wondering about the same with my own VA policy, rich in riders, which I fortunately purchased a few years ago. Although my particular policy and its unique inflation-adjusted income riders aren't available for sale anymore, I do have the option to add funds.
Now, I wasn't planning to invest more in the near future, but I wondered if perhaps I should accelerate my plans—before I completely lost the option. On the other hand, I wasn't comfortable having too much money in any one product and was concerned about subjecting myself to a new sequence of fees and surrender charges on new contributions.
On the one hand, on the other hand: so what to do?
I knew that the answer came down to probabilities, scenarios and risk tolerance, none of which I could do in my head. So, I gathered my notes and headed to the department of mathematics at York University, home to my esteemed colleagues and long-time research collaborators Thomas Salisbury and Huaxiong Huang. There, amongst the blackboards and the chalk, I described my dilemma and we boiled down the issue to a question of pure mathematics: "Should I add?"
Well, spoiler alert here. The answer was not quite what I expected. After a few weeks of computer simulations, intense number crunching and much back-and-forth, their answer was pretty blunt: "No, in fact, you should subtract."
Alas, the models unequivocally told us that if you own a GLB that is too good to last, then you should actually turn on the income as soon as possible. Exercise the option, initiate the rider and try very hard to "ruin" the account sooner than later. "Ah, but what about the roll-ups?" you wonder. Well, it seems they are red herrings. They are not big enough.
Excuse Me?
The mathematics can get hairy, but here is the nutshell version without the Greek. First, allow me to refresh your memory on the raison d'etre of the GLB. You invest money for the long term in a diversified portfolio of stocks and bonds, with the added benefit of not worrying about any income taxes (or fiscal cliffs) along the way. But, you pay 1-3% extra for a unique insurance policy (rider), which—to put it bluntly—gives you the right to 'crash' your account and still get income for life.
The rider is worth something—and worth paying for—because of the non-trivial probability that (a) the investment account might be depleted by withdrawals at some date, and (b) the individual annuitant might live well beyond that date. It's these two factors combined that create the value proposition to the client.
Importantly, your insurance (premium) payments come to an abrupt end if and when the VA account value actually hits zero. So, for starters, the sooner that account can be "ruined" and these insurance fees can be stopped, the better it is for the policyholder. Starting withdrawals sooner than later is a big help in this matter.
Here is the crux of it. The allure of a higher guaranteed base—if you wait a few more years before turning on the income—doesn't offset the value that comes from depleting the account as soon as possible. Remember that as soon as the account is ruined, the annuitant is now living off the insurance company's dime. And, while it might seem odd that trying to increase the cost to the insurance company is in the best interest of the policyholder, our rather extensive utility-based analysis indicates a similar phenomenon. You have to think like a game theorist. They lose. You win.
But it's more than trying to get your money's worth from the insurance company. Even if you (theoretically) pay absolutely nothing in fees you still shouldn't be enticed by the bonuses, roll-ups and ratchets. Yes, they might tempt you to wait, but our main argument is that they are not tempting enough.
Birds in Hand vs. Bush
Here is an example to help with the intuition. Your client is 65 years old. If he turns on the GLB today, he would be entitled to $10,000 per month for the rest of his life, plus the account value. The guaranteed lifetime income might increase if markets do (very) well, but it can never decline. 
So, assuming they don't initiate the GLB this year, how much more monthly income should they demand starting at age 66, to compensate for the forgone income? This isn't a question of personal preferences. It's purely in the domain of actuarial finance.
If you do the math, the breakeven delayed income would have to be $10,680 starting at age 66. You need an extra $680 per month for the rest of your life, which is an extra $8,160 per year, to make up for the lost $120,000 between age 65 and 66.
