AI pension advice

dunstonh

Long winded I know but I guess my question is how reliable is AI for advice?

I'm pretty heavily using AI. It's great for creating spreadsheets and doing some automations providing your instructions are thorough.

I sometimes throw questions at it just to check my thinking on areas where I want to reinforce what I'm doing. It frequently gives incorrect information. I then need to tell it it's wrong, and then you get the usual "whoopsie daisy" style message.

The thing with AI is it's right probably around 80% of the time. And it gives very convincing answers. The problem is you need to know what the right answers are and the wrong answers are.

In your scenario, I don't think the advice is very good. Obviously there's not a lot to go on, but 18 months in cash and 18 months in bonds and everything else in equities seems very high risk and unsuitable for the vast majority of UK consumers.

Another thing to be aware of is LLMs are not all created equal. I pay for Claude, ChatGPT, and Perplexity. I have Gemini within our Google Business software package. I also have copilot as part of Microsoft, who gave it away free for two years.

Co-pilot I find the weakest and I certainly won't subscribe to it in its current state when the expiry comes up.

Claude is the best at report writing. However, my instructions to get those reports would cover several pages of printed A4. Not something that you would generate in a chat box.

Claude is also better at spreadsheets, or rather was. I recently rebuilt one of my spreadsheets, and I started with ChatGPT. I was initially impressed, but it soon fell over. I then switch to Claude which made a much better job of it but in the last few months, they've really hit credit use hard.

Gemini, though, on Google Sheets in its current beta is proving to be very good. However, Gemini is very prone to hallucinations when asking it questions about subjects.

Perplexity is a great replacement for Google. However, I fear its longevity. Like Claude, they've been hitting usage credits hard and whilst it is still a very good Google replacement, silence it's increasingly losing its usefulness for everyday tasks unless you pump a lot of money into it.

The biggest problem with all the AI LLMs is that they will tell you what they think you want to hear. They will say it very confidently. And it's very easy to believe them when they're wrong.

The obvious drawback was in a bad year or 2 that 3k a month turns into 2 k a month and my 50k emergency fund gets utilized.

What about lost decades?

If you'd gone 100% equities on the 1st of January 2000, ten years later, your value would be lower without any withdrawals. With withdrawals, you'd be close to running out depending on your draw rate.

DRS1

https://forums.moneysavingexpert.com/discussion/comment/82058622#Comment_82058622

Yes the plan is something which is often talked about here as a drawdown strategy. Cash to cover x years bonds (ILGs probably) to cover the next x years and equity investments to cover the rest and hopefully grow but also have time to recover if they fall in the near future. But the question is whether such a strategy can support the level of income you want with the assets you have.

DRS1

https://forums.moneysavingexpert.com/discussion/comment/82058701#Comment_82058701

Yes You have said that. What you haven't said is what relationship those figures bear to what you actually spend. Of course you don't need to say that but do show you understand the question. Think Charles Dickens

"Annual income twenty pounds, annual expenditure nineteen nineteen and six , result happiness.
Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery”

The decreasing income idea is not a bad one (I'd go along with the get it while you can enjoy it line) but it does ignore inflation. Essential expenditure is unlikely to fall. It is only the fun bits that you are likely to drop as you get older (and then they may be replaced by the dreaded nursing home fees which will eat through your 400k in next to no time).

Mozza001

The question above could such a strategy support the level of income, again who knows anyone on here, my IFA anyone at all is guessing. I'm pretty sure there can be built in financial plans/safeguards to such a plan, the obvious one being slash the withdrawals during bad times wtc

Mozza001

What about lost decades?

If you'd gone 100% equities on the 1st of January 2000, ten years later, your value would be lower without any withdrawals. With withdrawals, you'd be close to running out depending on your draw rate.

How often in history has this happened, how likely is it to happen again?

I get the caution, but if you look at things with very worse case scenarios all the time people will be working until 80.

Mozza001

I have a hyperthetical Dunstonh if I may?

If you had a client 100% in equities at his request at the start of this "lost decade" by 2010, with your advice would his initial investment still not have recovered, ie would you have advised he stay as he was for 10 years?

I'm curious how investors coped at this time, whose money was managed by professionals

dunstonh

https://forums.moneysavingexpert.com/discussion/comment/82058728#Comment_82058728

The IFA will have software that does modelling. Whilst modelling is based on historical data it can at least point you in the right direction. For example, if your IFA is telling you that your current planning is only going to give you a 20% success rate of not running out of money over your lifetime then you pretty much know that you've got to put more money aside, extend your working life or take less in retirement.

I'm pretty sure there can be built in financial plans/safeguards to such a plan, the obvious one being slash the withdrawals during bad times wtc

That assumes you have an excess in discretionary spending that can be reduced.

How often in history has this happened, how likely is it to happen again?

Enough times that you should be factoring it in as a high chance of happening during a 30 to 40-year term.

A sensible cash float would have reduce the impact significantly. Also, having greater diversification into alternative asset classes would have helped.

I get the caution, but if you look at things with very worse case scenarios all the time people will be working until 80.

Planning on the very worst-case scenario is probably unnecessary. Statistically, it's highly unlikely to happen.

If you had a client 100% in equities at his request at the start of this "lost decade" by 2010, with your advice would his initial investment still not have recovered, ie would you have advised he stay as he was for 10 years?

Depends if they were in the accumulation stage or the de-accumulation stage. I had nobody at 100% equities in the decumulation stage. I doubt I ever will.

Those in the accumulation stage, still contributing into their investments on a regular basis, then lost periods are good news rather than bad news.

I'm curious how investors coped at this time, whose money was managed by professionals

you generally find that advised portfolios are more cautious than DIY investors and YouTube videos. especially as many of the latter only look at recency rather than the long-term

Albermarle

https://forums.moneysavingexpert.com/discussion/comment/82058728#Comment_82058728

Surprised that nobody so far has mentioned Safe Withdrawal rates.

Based on historical statistical modelling a theoretical safe withdrawal rate can be calculated, where there is only a 5% chance of running out of money by the time you are 90/95. At your age this would be around 3.5%, meaning you would start with about £1150 a month , which would increase with inflation each year.

If you were able to reduce expenditure in a market downturn below £1150 ( particularly if this happened in the early years ) and increase it in an upturn, you can in theory at least increase the average withdrawal rate to say 5% - £1,700 a month.

You could increase the risk of running out by taking some more, but if you go too far and markets are not kind, the pot could run down scarily quickly .

Of course as mentioned the state pension will help, so you could also take more before age 67 and less afterwards.

However overall the AI plan looks too optimistic, and the risk of taking the level of income mentioned is for sure more than 'moderate'

Mozza001

Ran some scenarios again last night.

AI must be broken, or winding me up

See what my IFA says when he stops laughing in August.

barnstar2077

https://forums.moneysavingexpert.com/discussion/comment/82059165#Comment_82059165

I was surprised too, especially as I mentioned it earlier in the thread! (or maybe I talked around it, it is open to interpretation I guess! : )

For the OP: My plan involves hitting a target number by age 53/55 (currently 53 and £400k, consisting of £100k ISA, £300k pensions), then ignoring inflation and growth, and just dividing my money into pots/buckets.

x amount to get me to 57 from my ISA (amount changes depending on when I stop work, but £56k from 53 to 57 for example), then ten years worth of money from my pensions to get me to 67/68 and state pension age (£160k is current thought.) Then taking a few grand (or possibly an annuity by that point) from what is left in the pensions to top up SP. Whatever is left in ISA after 57 is house maintenance and white goods fund.

So, £56k + £160k would leave me with £44k maintenance fund in ISA and £140k in pensions to top up SP. This plan also has a lot of slack in it, so that no matter what I should at least make it to SP age before my money runs out (worst case scenario.)

In reality I will spend more or less (especially during the ten years before SP) depending on how returns and inflation are going, but I have to start from somewhere. Even if my assets fell 50% the day I retire I will still be fine. I just feel like time is running out and I want to make the most of the 30 summers I will have left on average from age 53!