How to trade the Unholy Grails system?

[InsvestoBoy]

The universe was BT Margin allowed margin traded stocks.

Thanks @tech/a can you confirm my whether my description of how your margin account worked is correct?

i.e. if you had $30k you were able to own up to $90k in stock and if you managed to turn a $30k profit, you'd have a $60k balance and be able to own up to $180k in stock?

Also can you help us understand how tax worked on your trading system? As you traded it for 7 consecutive years you were generating net capital gains and must have been required to pay CGT if trading in your own account or corporate profits if you had setup a corporation to trade through. But the equity curve doesn't seem to show any impact of annual or quarterly tax payouts on capital gains and compounds regardless.

How could you achieve this, unless you were paying tax from another source of funds?

 

[rnr]

↑
Thanks @tech/a can you confirm .....

Also can you help us understand how tax worked on your trading system? As you traded it for 7 consecutive years you were generating net capital gains and must have been required to pay CGT if trading in your own account or corporate profits if you had setup a corporation to trade through. But the equity curve doesn't seem to show any impact of annual or quarterly tax payouts on capital gains and compounds regardless.

Hi InvestoBoy,

I think it would be fair to say that the results of most system tests posted on a forum are on the premise that surplus profits are reinvested in the system. Having said that, I believe I have seen a couple of instances where a set amount has been taken out annually from profits.

Cheers,
Rob

 

[InsvestoBoy]

↑
Hi InvestoBoy,

I think it would be fair to say that the results of most system tests posted on a forum are on the premise that surplus profits are reinvested in the system. Having said that, I believe I have seen a couple of instances where a set amount has been taken out annually from profits.

Cheers,
Rob

Hi rnr, the results in question are not from a test they are supposed to be live trading results...

But even for backtesting purposes, I'd posit that if people aren't making some accountancy for paying something akin to the corporate tax rate of ~30% on your FY profits, they can't really rely on metrics like equity curve, CAGR, Sharpe produced by the backtest as meaningful because the premise that you can compound all profits is just plain wrong. Obviously some ratios like win/loss and expectancy are unaffected.

 

[tech/a]

Re margin yes you could but that’s not how it was used

Re tax
Profits were bulked together with my own tax as Director of RWI
Along with a lot of property holdings and sales
I never drew on the account to pay tax
I have an accountant for the company and an accountancy firm for personal and company
They work their magic.

Took all the funds remaining to buy property
There was quite a lot left that I’d held for over 12 mths

 

[rnr]

@InsvestoBoy

But even for backtesting purposes, I'd posit that if people aren't making some accountancy for paying something akin to the corporate tax rate of ~30% on your FY profits, they can't really rely on metrics like equity curve, CAGR, Sharpe produced by the backtest as meaningful because the premise that you can compound all profits is just plain wrong. Obviously some ratios like win/loss and expectancy are unaffected.

Agree with your comments on this one.

 

[InsvestoBoy]

Re margin yes you could but that’s not how it was used

Please tell us for the posterity of this thread, and so we can make a useful comparison versus a benchmark, how it was used.

 

[willy1111]

Thanks @tech/a
Also can you help us understand how tax worked on your trading system? As you traded it for 7 consecutive years you were generating net capital gains and must have been required to pay CGT if trading in your own account or corporate profits if you had setup a corporation to trade through. But the equity curve doesn't seem to show any impact of annual or quarterly tax payouts on capital gains and compounds regardless.

How could you achieve this, unless you were paying tax from another source of funds?

From my point of view... all funds available to the public, indexes etc show returns before tax, so from a comparison point of view I think it is better to show returns like this.

A VAMI (value added monthly index) spreadsheet allows one to do this.

Simply compare the account balance (market value of portfolio plus cash) at the start and end of the month, allowing for additions and withdrawals (tax) to show a monthly return which is then applied to an arbitary figure say $100k. Can also do it with a benchmark like All Ords Accumulation Index to compare returns to the benchmark.

If one wanted to know the before and after tax return simply make an additional withdrawal column just for tax when it is withdrawn from account.

 

[qldfrog]

But it is true your results in Australia can not really be that fantastic with our tax rates, or worse the cgt which is not inflation indexed..what a joke

 

[tech/a]

Please tell us for the posterity of this thread, and so we can make a useful comparison versus a benchmark, how it was used.

Initially I funded enough to trade 10 positions risking 10% on any one trade.
This was around $34000 my own capital and the rest Margin.

As equity grew and I closed winning and losing positions the account eventually became self funding.
Each new position was the same $10000 with $1000 risked.

Still using the BT margin list.
It was my opinion that BT had done their research and wouldn't allow a dud in that list!

Margin was only used to get going.
The idea was to show how a small account could profit in the longer term.
Often The portfolio was full of its 10 positions some at $10,000 and others deep in profit
a great deal more in the end the liquidated value had some very profitable stocks in it
I remember 4 being over 10X initial purchase.

From the top of my head QBE was one. Think TAH was another. That pulled over 10X initial capital over the life of the portfolio.

 

[InsvestoBoy]

From my point of view... all funds available to the public, indexes etc show returns before tax, so from a comparison point of view I think it is better to show returns like this.

Not sure how that would be a valid comparison? If you invest in AFI, or VAS, sure they show the returns before tax, because you can hold them indefinitely without selling and causing a CGT event to occur. The systems described in Unholy Grails are constantly generating net capital gains and causing CGT events. The compounding in the equity curves and reported Sharpe ratios etc is only valid if you can pay that tax from another source. If you have to pay CGT or corp profits tax out of your trading profits then they are obviously overstated.

 

[InsvestoBoy]

Initially I funded enough to trade 10 positions risking 10% on any one trade.
This was around $34000 my own capital and the rest Margin.

As equity grew and I closed winning and losing positions the account eventually became self funding.
Each new position was the same $10000 with $1000 risked.

Thanks @tech/a, so it would be correct to say that after your run had generated ~60k in closed profits, you were no longer using the margin loan?

Still using the BT margin list.
It was my opinion that BT had done their research and wouldn't allow a dud in that list!

One idea I had recently was to build a universe from all of the fundamental managers who list their ASX holdings, that way you can essentially leverage the in-house fundamental research of many different firms.

 

[tech/a]

Don't think they are over stated.
If your wage before Tax is $100K your not overstating your wage
by having to pay tax from it.
You earned it just as a portfolio earned it.

 

[InsvestoBoy]

Don't think they are over stated.
If your wage before Tax is $100K your not overstating your wage
by having to pay tax from it.
You earned it just as a portfolio earned it.

@tech/a I am specifically referring to the compounding portion of the curve.

If you start trading an account with $100,000 and assume you can generate 10% per annum returns, in year 1 you will have $110,000.

If you pay no tax out of that and compound from there you'd have $121,000 by year 2.

Instead if you have to pay 30% CGT on that $10,000 then you'd only have $107,000 to compound from.

Generating 10% on $107,000 would net you $10,700 and after paying 30% CGT on that amount you'd have $114,490... 5.6% less money to compound with in year 3.

Those differences would add up to quite a lot over 10 years of trading versus the backtested equity curve.

If you're paying corporate profits tax quarterly the compounding equation would be quite different again.

Surely this is obvious and I'm not missing something...

 

[tech/a]

Thanks @tech/a, so it would be correct to say that after your run had generated ~60k in closed profits, you were no longer using the margin loan?

Yes.

 

[tech/a]

@tech/a I am specifically referring to the compounding portion of the curve.

If you start trading an account with $100,000 and assume you can generate 10% per annum returns, in year 1 you will have $110,000.

If you pay no tax out of that and compound from there you'd have $121,000 by year 2.

Instead if you have to pay 30% CGT on that $10,000 then you'd only have $107,000 to compound from.

Generating 10% on $107,000 would net you $10,700 and after paying 30% CGT on that amount you'd have $114,490... 5.6% less money to compound with in year 3.

Those differences would add up to quite a lot over 10 years of trading versus the backtested equity curve.

If you're paying corporate profits tax quarterly the compounding equation would be quite different again.

Surely this is obvious and I'm not missing something...

Yes again unless you pay tax from some other source.
The curve and compounding component is dependent on not eating into profits.

 

[InsvestoBoy]

The compounding in the equity curves and reported Sharpe ratios etc is only valid if you can pay that tax from another source. If you have to pay CGT or corp profits tax out of your trading profits then they are obviously overstated.

Don't think they are over stated.

Yes again unless you pay tax from some other source.
The curve and compounding component is dependent on not eating into profits.

That's what I said heh!

All the equity curves in Unholy Grails and the one you posted for TechTrader should really come with a big disclaimer:

Only valid if you are rich enough to pay taxes from other funds!

 

[tech/a]

That's what I said heh!

All the equity curves in Unholy Grails and the one you posted for TechTrader should really come with a big disclaimer:

Only valid if you are rich enough to pay taxes from other funds!

Money makes Money!

As should every test I think I've ever seen.
Including Pete's portfolios and every other portfolio demo I've seen.

Your perfectly correct and those who wish to trade for a living are way off the mark in their calculations of capital needed to start their trading business.
From living expenses to tax.

 

[willy1111]

Not sure how that would be a valid comparison? If you invest in AFI, or VAS, sure they show the returns before tax, because you can hold them indefinitely without selling and causing a CGT event to occur. The systems described in Unholy Grails are constantly generating net capital gains and causing CGT events. The compounding in the equity curves and reported Sharpe ratios etc is only valid if you can pay that tax from another source. If you have to pay CGT or corp profits tax out of your trading profits then they are obviously overstated.

We are all different.

From my point of view, it is a valid comparison as I wish to know what the Compound Annual Growth Rate (CAGR) is of AFI or VAS or any other system before tax.

Why? because then I can gauge the annual return expected from a given amount of Capital, ie VAS may have a CAGR of 8.34% for the last 10 years compared to a system with a CAGR of 20% for last 10 years.

If I have $1M capital to allocate - I'd expect roughly a return of $83,400 from VAS and $200,000 from the system - I can then deduct the necessary tax depending on applicable tax rates at the time, it might be Pension mode in SMSF and no tax payable. The net after tax is then available to me if I wish to pay for living expenses or re-invest.

So I find it helpful from that point of view.

Perhaps you are approaching it from the view of if I have $100K now, how many years compounding at x% before I have $1M in capital and not including the tax drag in x% will impact the time.

 

[rnr]

Hi InsvestoBoy,

Pressured for time yesterday arvo whilst responding to your post so had to cut my response short.

When evaluating which system to use one should factor in the impact of income tax on each system.
The very nature of the system, such as trend following in tech/a's case, will also have an impact on the timing of a taxable profit verses a system profit and hence the timing for payment of any income tax, whereas with a mean reversion system the system profit and taxable profit are likely to be almost the same.
The other issue that needs to be determined is the entity used to run the system (eg. individual, partnership, company, discretionary trust, unit trust, SMSF) as this will determine the rate of tax payable as well as initial setup costs and ongoing annual running expenses .

It is important that anyone involved with system development understand the impact of the above on their system metrics.

Cheers,
Rob

 

[jjbinks]

So I backtested the most straight forward 3 systems from unholy grails on US and AUS Markets.
(Bollinger Band, 100 day high, MA Cross)
Results are from orginal time period mentioned in the book. And more recent time.