the penny and the steamroller at Monte Carlo

Note: the lower part of this post has rapidly moving/flashing graphics

I am slowly inching my way to being able to run my DC pension flat in front of my DB pension, effectively burning up the DC fund over five years, drawing my DB pension at normal retirement age. The latter will give me enough to live on, and I can keep my ISA tax-free income reinvested against the day when wages in Britain begin to increase in real terms, whereupon I will start to fall behind without being on the side of Capital. Or I can use it to hedge against the demise of the NHS, when healthcare will become at best insurance-based like in most other European countries, or at worst like the evil that is the US system. I am fortunate enough to avoid doctors and hospitals at the moment, but nobody gets healthier as they get older.

Running a DC pension flat over five years is an odd time-frame

There’s an age-old rule of thumb, never have money you will need in five years anywhere near the stock market. Ever ask yourself why this is so? The basic reason is that the equity premium is tiny – about 1/20th of the capital sum per annum, and this is the fella who is fighting on your side. He is such a puny bastard you either need a lot of time or a lot of money on your side. They ain’t making any more time, which is why compound interest won’t save your ass unless you work till you have one foot in the grave…

On the other side of the ring is the Bad Guy – volatility. He’s a moody bastard – sometimes euphoric, and sometimes down in the dumps. He will drag the aggregate value of your share portfolio up and down as he pleases, and that volatility is massive, because Mr Market is the 600lb gorilla, and if he’s in a grump everybody gets to know.

So I asked myself what would happen if I started out with a lump of about 80k – if I want to run it flat in five years I can’t really use much more than that, and there is no point in running my DC SIPP into the revenue from my main pension as that is already over the personal allowance, so I’d be taxed on the SIPP in its entirety.

Right off the bat I can take ~£20,000 as a tax-free lump sum, and this is the obvious thing to do. [ref]the PCLS is a special case, the actual income from the SIPP is the tax-free personal allowance each year[/ref].

After taking the PCLS I can take ~£11,000 per annum income tax-free from the personal allowance, and simple arithmetic shows that after I have iced £20k from my stake I have £60k to run out, and 5×£11,000[ref]the personal allowance is £10,600 at the moment[/ref] is 5k short of the £60k balance. If I am lucky they will lift the personal allowance towards the end of this Parliament[ref]the aspiration that nobody on the full-time national minimum wage should pay tax ought to help with that[/ref] and an Ermine will squeak under the finish line with the goose feathers unplucked.

What about that steamroller then?

I have paid a lot of tax in my time, and I considered contributing £3600 a year to the SIPP over those five years at a cost of £2800 p.a.. Somehow I want a way for the taxman to share in the risk, so I asked myself the question of what I think about tackling the steamroller of the stock market

What would happen in my SIPP if rather than cash I held it in some index fund of a market that isn’t on a high CAPE, so FTSE 100 or 250 rather than S&P500…

I tried a Monte Carlo simulation. As an engineer I am deeply opposed to using statistical analyses without knowing their limitations. Stock markets are a bastard because they do not obey the central limit theorem, there is serious correlation between the supposedly independent actors and as such market distributions are fat tailed – shit happens (and euphorias happen) more often than simplistic risk calculations assuming normality tell us. Taleb wrote a book called Fooled by Randomness riffing on just that topic. So bear in mind that what you are about to see is an optimistic simplification on the volatility. And nobody knows exactly how optimistic! That steamroller is big, and it is virtually guaranteed that there will be a stock market crash in the next five years IMO.

I ran 50 Monte Carlo trials each containing five independent scenarios, 250 trials in all, a fifth shown spatially (like FireCalc) and 80% temporally (like nothing else I am aware of) assuming that each year I take out a personal allowance-worth from my portfolio, from that money I add £2880 to which the taxman adds £720.

The cash scenario assumes I get -1% interest on cash in real terms due to inflation. I have assumed an equity premium of 4.3% real, fees of 0.45% (HL SIPP, funds, no transaction charges) and a FTSE100 annual standard deviation  of 30% (about 2% rattle a day × √256 trading days p.a.). This is probably very harsh on the FTSE100, it doesn’t normally vary by 2% a day, so I will simulate more dramatic spreads than is probably the general case. I derived that from looking at the values now when there seems to be some upset in the air. As a prospective purchaser, this pleases me, but I should do more work to get a figure more characteristic of a five-year period.

Immediately this tells me that five years of contributing £3600 will leave me with a rump of £20000 on a cash only basis that I’d have to pay tax on[ref]this is offset by the fact that I will have claimed 20% tax on it on the way in, tax I will never have paid, but is perfectly legit to reclaim in a SIPP as a non-taxpayer. The SIPP provider does it for you, you don’t have to do anything[/ref]. This is not earth-shattering, but I wondered if I could use the taxman to underwrite some of my stock-market risk – I have the choice of whether to put in that £3600 p.a, which makes the taxman contribute up to £3600 across the five years.

As you can easily see from the variable parts, taking stock-market risk is distinctly favourable on average – I am happy to pay tax on money if it’s more than the cash position – easy come, easy go.The taxman’s bite gets less if I get hurt by the stock market, and to some extent that makes me wonder if I should take a bit more risk on his dime. Maybe I should only put in the non-earner contribution if I am taking flack.

If I pump up my equity exposure to 60%, a common value for young ‘uns saving for 30 years[ref]If I were such a young ‘un 30 years off retirement I’d be 100% equities – let’s face it I am considering that young whippersnapper’s split as an old git five years off running the lot into the ground![/ref] there are a significant number of scenarios where I run out of money. Now that’s not unreasonable – remember that I am drawing down at a racy 20% p.a. rather than a steady as she goes 4%. I’m not actually going out and buying a Lamborghini, but my aim is to drive this into the ground ahead of my DB pension in five years. So of course I am taking the piss drawing down so fast.

But look at the potential for upside gain (with the volatility possibly overstated, I will study how to do this less inaccurately). I could take the line that forestalling running out of money at 60% split is what half my PCLS is for. Alternatively I should only contribute the £3600 if I am being clobbered in the markets, because I can get the taxman to help me with 20% of the downside risk. If I am doing well I should keep the £2880 and invest it unwrapped or in my ISA if there’s space. To simulate that properly I have to run this as a Excel VB function, it’s too mind-numbing and untestable to do all that conditionality with cell formulae and IF statements. But it is possible to specify this as a mechanical strategy.

Wealth warning:

if you are in the accumulation phase, do not extrapolate from this that you will have such a rough ride. Remember I am drawing down at 20% – as a result if next year is a bad one on the markets and I don’t slow my drawdown rate I will impair my performance simply by selling 11k of units in a market rout. You won’t be doing that, so those units live to fight another day!

I have represented the taxman’s bite at the end by multiplying the terminal amount by 0.8, because I have to pay tax on the entire amount I have left over after 5 years, because I will then draw my main pension, which is greater than the personal allowance. Obviously if I’ve flattened myself there’s no tax to pay 😉 I’m not terribly worried about getting steamrollered in this SIPP. I saved this money for a different purpose, but Osborne’s changes let me front-run my DB pension with it, effectively investing it in increasing my defined benefit pension, as opposed to my initial plan of drawing the pension early, taking an actuarial reduction and investing the AVC. I have enough defined contribution savings in the form of my ISA, although half the PCLS will probably end up in there at some stage, with the rest to follow if I don’t get slaughtered in the SIPP.

I was surprised by:

• the fact that the odds in favour of the stock market were clear for such a short timescale
• the sheer variability of sequence of returns risk – the integration time of five years is not enough to start to bring these together (if that will ever happen – firecalc seems to indicate a bad start isn’t cancelled over 30 years, and it may be that the high drawdown inherently changes the statistics). The moving presentation kind of brings that sequence of returns risk home in a more visceral way to me than the straight firecalc trace 😉
• the worst outcomes all seem to start off getting hammered by the market in the next year. As indicated in the wealth warning, that’s not terribly surprising, given my high drawdown rate. I would be wise to use the cash from the 40% cash part to avoid being a forced bear market seller, bear markets often come back materially within a year. The simulation assumes I draw down from cash and equities in proportion to the holdings. I am not sure I would do that in practice – I’d first draw the cash flat.
• The opportunity to only put in the £3600 p.a. if I get hammered in the early years gives me a way to get the taxman to sponsor some downside risk

The cash conundrum on the PCLS

The biggest hazards in the five year term that I can see is serious inflation, and also a serious financial crisis.  Both of which resulting from a fallout of the tricks used to kick the last one into the long grass. Cash troubles me generally – I really can’t be arsed to yomp a few grand from pillar to post to try and win a bit more interest. I am not constitutionally careful enough to line all the ducks up in a row, and I didn’t give up work to stare at screens to try and rake 5% p.a. on about 20k that seems to be the practical limit. £1000 a year is nice, but I prefer to put my time into trying to understand markets, risks and my financial situation better rather than to slog minimum wage opening bank accounts and fretting if all the direct debits go out right. There’s n’owt wrong with that if it’s your bag, but it just isn’t how I want to spend my days. Hell, I’d even consider working if I needed £1000 that much rather than fart about that way 😉

I am thinking of holding some (about 5k) with Zopa, but unlike most PF writers I view Zopa as stupendously high risk, higher inherent turkey distribution risk than a stock market index fund. The risk becomes obvious when you take a look at what Zopa borrowers are borrowing for. I look at these dudes and I see far too many people borrowing for cars

Total Amount Borrowed: £10,000

Age: 32

Purpose: Car

Location:Teeside

£10,000 borrowed for a car, FFS! I have never spent £10,000 on a  car in my life, and if you are spending that much a car then you should be rich enough to pay upfront… Here we have another paragon of fiscal probity, borrowing to repay loans

Total Amount Borrowed: £15,000

Age:40

Purpose: Consolidate existing debts

Location: Manchester

This is a car crash in the making, those debts ain’t ever gonna get repaid…

All these iffy borrowers will be able to service their loans, right up until we have the next financial crisis which will happen some time in the five years I am thinking of, and above all else when interest rates go up, because most everybody in Britain is overexposed to that financial weapon of mass destruction, the housing market, a mistress that whispers sweet nothings and dreams of endless riches in the good times, and then throws her lovers out on the street naked in the middle of the night when she has a hissy fit. I know, I saw it happen to my neighbours in the 1990s, and it’ll happen again. When your kids are wailing because you haven’t got a roof over your head, Zopa lenders can whistle in the wind for the money they lent you. I think there’s a significant risk of this happening in the next 5 years.

Against this the stock market is at least down a decent amount on those stupid levels of 7000 plus it was earlier this year, and while the index may tank it won’t be wiped out. But that variability is counter-intuitively huge, so I will study the results of profiling it, varying the cash/market balance.  I have the privilege that I can take a reasonable level of risk with this, if I am wiped out in three years then I will have still done okay on the 40% tax I didn’t pay earning the money and the market gains from 2009-2012. Nevertheless taking a flyer on a 10% risk of losing half the value (I will still pay out 30k of the drawdown as income in the worst case scenario of the simulation, though  the risk is a bit higher due to those fat tails) is kind of straight between the eyes and not for the nervous of disposition.