Multi Manager Funds - Hargreaves Lansdown

Dooooooooooooooonut

Wilmott is to quants and 'new school' traders (not many locals from the LIFFE floor there, put it that way) the same way that moneysavingexpert is to people who like saving money.

Paul Wilmott is a great bloke, everyone will tell you this.

purch

Is the Wilmot site anygood?

If you can answer this question,

Is Dupire formula the right way for dealing with absolute dividend paying asset local volatility ?

then it's the site for you !!!!

If like me, you think 'Volatility' is a way of measuring how angry Aegis get's when he answers Dooooooooonuts posts, then maybe you should stick here !!!

Dooooooooooooooonut

Of course that highlights an important point, just how does the typical investor get their head around volatility? It's not so long ago that, if modelled at all, it was modelled one dimensionally... and that's about as far as any !!!!ing !!!!!!!! 'key facts' or 'fund prospectus' will go... and that is nowhere near the true story - actually we don't have a satisfactory volatility model at all (expect possibly for commodities), hence poor saps just not getting it.

Medium risk !!!!!!. What does that even mean?

fimonkey

oooh dear, I think I'm a poor sap!.. ah well, I can cope with that and see that I should continue to hang out on this site then.

Doonut, your theories suond fascinating to me, but beyond me too. Anywhere where their general principles are explained in 'poor sap' language please?

Cheers again.

purch

Quantative Finance is the "science" :eek: of pricing and hedging derivative securities and managing risk.

Options, dynamic hedging, interest-rate modeling, portfolio theory, price forecasting using statistical methods, etc. are all part of this.

It's for clever Math 'geeks'.................the types who invented CDO's and all the other wonderful stuff that bankrupted the Banking system when put in the hands of people who didn't understand it properly.

fimonkey

ah, I might not understand the details, but I see an overall picture emerging and it kinda goes aganist one of my mantra's which is:

"Make everything as simple as possible,- but not simpler".
Albert Einstein

Dooooooooooooooonut

fimonkey wrote: »

oooh dear, I think I'm a poor sap!.. ah well, I can cope with that and see that I should continue to hang out on this site then.

Doonut, your theories suond fascinating to me, but beyond me too. Anywhere where their general principles are explained in 'poor sap' language please?

Cheers again.

sorry this took a bit for a reply, fixed income went haywire this afternoon... And then I did good and went to pub but I don't think it will affect my writing... doing this from a phone in fact.

Volatility... hehe, where to begin

Let's start off with a really informal definition of volatility: "how much we can expect the price of a security to move around"? Why do you need this? Simple: You need to know how much you can lose or make from a trade, particularly if you are leveraged in any fashion. I assume you can see why this important when you are buying a fund: in short volatility is a guide to risk. In fact by my definition voltaility IS risk and reward, but this is at odds with everything an IFA (or to be fair any finance book) will tell you: MISTAKE NUMBER ONE.

So the question comes about how to measure volatility... one idea that would come to mind is the average of how much the security moves in a day... wrong again. As simpleton aegis will tell you, the sum of INDEPENDENT (<== MISTAKE NUMBER TWO) RANDOM (<== MISTAKE NUMBER THREE) events will eventually come to a normal distribution (TRUE BUT GIVEN TWO AND THREE THIS IS GOING TO HAVE MASSIVE CONSEQUENCES FOR THE FUTURE), and when this was first being tried out (well the first time it was tried out was a chap who you will find named in your purple question, but ignore him as he was French and the 'academics' so prized by Aegis didn't understand such and doomed his career) it appeared that stocks, certainly, followed a roughly normal distribution in terms of their price changes from day to day, week to week, hour to hour... that is to say they were moving as though in brownian motion. So what's the obvious measure of volatility here? Well we said they moved along a normal distribution (and of course the sums of normal distributions are more normal distributions)... so we could measure volatility by standard deviation. (AGGHGHGHHGHGHG THIS WILL LOSE SO MANY PEOPLE SO MUCH MONEY IN THE FUTURE AGHGHGHGHGHG).

Result? Say we have obtained the standard deviation of an asset price (almost certainly by looking at past returns [<== MISTAKE NUMBER FOUR]). Say this is the standard deviation of the price over the year, I'll type SD... you can reasonably say that after one year the current price +/- 3SDs will be the price of the asset 99% of the time (this follows from the properties of the normal distribution).

Ok dooooooooooooooonut you've explained how volatilty can be used to measure risk so WAIT I SHALL CUT YOU OFF THERE. Sadly some of the same gentlemen (or at least all tied to Chicago, I really can't remember who 'rediscovered' asset prices following a normal distribution) were at this point working on how to price an option. Classically (or to be fair to them, this was their great insight) to price an option one thinks about how one will hedge it in the underlying; i.e. hedging delta. That is to say, a very out of the money option has very small delta so to hedge it we have a little bit of the underlying, an in the money option has delta of 1 (we need the full underlying) and everything else is in between. The black-scholes(and you can add merton if you want) formula that was designed to work out how much these options (essentially insurance contracts) comes about by considering how much it would cost to hedge these options by hedging delta, that is buying or selling the underlying as the price of the underlying changes.

Oh, and it assumes prices are normally distributed.

Oh, and options (of a kind; classic black scholes is related to european options) are involved in almost all structured products.

Oh, and prices are not normally distributed (AH HA! THIS OPTION PRICING WORK IS GOING TO ALMOST BRING DOWN THE FINANCIAL SYSTEM IN 1987 [DYNAMIC HEDGING], 1998 [VAR ETC STEMMING FROM NORMAL DISTRIBUTION]... AND THIS KIND OF MISTAKE DID A !!!!ING GOOD JOB OF IT OVER THE LAST YEAR AS WELL)

See, volatility is not static at all. It comes in clumps (there have been attempts to model this, as far as I know starting with Mandelbrot but then he wasn't from Chicago, he just had a clue) and outlying events happen far more often than the normal distribution would predict - AND THESE ARE THE EVENTS THAT MATTER.
Why does this happen? There are a myriad of reasons - just a few off the top of my head - So to quote Einstein as you like him (as do I)... he may have been wrong (was he? I dunno) about God playing dice with the universe but sure as !!!! God isn't playing dice with market prices. Human beings are setting those prices, and for whatever reason humans just won't follow a mathematical model. For a brilliant example of this take a looke at Madoff. He exhibited very low volatility for the returns... until he blew up (hilariously, mind you; I think he sounds a brilliant character) - try to predict that with their silly models. Or look at LTCM in 1998 - once people knew they were !!!!ed they got front run to !!!!!!y and all their silly correlations went to one. Good. And by the way this included a chap from said black-scholes formula Indeed, even if acting according precisely to previous returns you will find that these outliers happen far, far more often than normal distribution - this is the so called fat tails (if you were in the pub with me I would draw you adiagram to explain) and there have been attempts to model this through power law distributions as well but they STILL get it wrong - BECAUSE THIS STUFF IS FUNDAMENTALLY UNKNOWABLE. IT JUST HAPPENS ONE DAY.

So how does this affect you? Well, basically, when an IFA or whoever explains the risks of a fund to you one of the major things they are looking at is past (and presumed present) volatility. But they haven't a clue what they are doing. This is !!!!!!!!. So you have very little idea what you are getting into. The best you can do is try and REASON THROUGH IT yourself and take a view; and if you're not prepared to take a view about the risks yourself then you shouldn't invest in anything - and that even includes cash - so in other words you have to take a view, basically.

So how do I feel about volatility? My personal, philisophical view is that it cannot be measured at all - even if you have a working system (and, to draw something from past performance, this looks unlikely at best) for a millennia something or other will happen and it will have been of no use whatsoever

I felt so good typing this, as I genuinely felt it will be helpful, that I have delted my comments about IFAs and other urchins before posting it. I hope the facts should speak for themselves AND you will see why I find so much of their advice utterly, utterly useless. I've also trimmed it as much as possible to leave only the relevant bits in, actually there was a huge irrelevant ramble in the middle about hedging exotics that I don't remember writing and it was 20 minutes ago!

PS !!!! IFAS
PPS !!!! ACADEMICS APART FROM THAT JOHN COATES GUY WHO HAS NOW VALIDATED MY RING FINGER OR SOMETHING I THINK
PPPS IF YOU WANT FUTHER CLARIFACTION OF ANY POINTS AM HAPPY TO GIVE IT BUT MAY TAKE A WHILE

Lokolo

Dooooooooooooooonut wrote: »

So how does this affect you? Well, basically, when an IFA or whoever explains the risks of a fund to you one of the major things they are looking at is past (and presumed present) volatility. But they haven't a clue what they are doing. This is !!!!!!!!. So you have very little idea what you are getting into. The best you can do is try and REASON THROUGH IT yourself and take a view; and if you're not prepared to take a view about the risks yourself then you shouldn't invest in anything - and that even includes cash - so in other words you have to take a view, basically.

Thing is, it is a good way of judging risk, because if its very volitile (spelt right? ) then people start to get jitters when it goes down 20%, up 35% etc. and I am sure anyone who doesn't know about investments is going to think this.

So if it starts going crazy over a month, the investor (who knows nothing) will start going, actually this is too risky for me and wants to pull out.

And thats just common sense, not any maths or normal distribution involved. If someone doesn't like risk, volity is something thats not going to be good!

And yes I see that it cannot be predicted or anything, but you can usually figure out that if a fund has been very volitile for 20 years its doubtful it will suddenly stop. Would a racing horse who comes last in 8 years running start coming 1st in every race for years on end? No, don't be silly lol.

Dooooooooooooooonut

But you are assuming that volatility is constant (i.e. one dimensional) as I alludided too o earlier. It isn't. The safest things (measured by past volatility) can all of a certain go all buckaroo!

Agreeed something very volatile is very unlikely to go calm [unelss it goes to 0 hehe], but if it does, then who cares? It's the opposite that is far more dangerous AND far more likely!

take a look at a chart of the vix

whiteflag_3

Dooooooooooooooonut wrote: »

PPPS IF YOU WANT FUTHER CLARIFACTION OF ANY POINTS AM HAPPY TO GIVE IT BUT MAY TAKE A WHILE

Sorry Deemy, can you just explain that stuff about volatility again, not sure I understand.?