Index | Recent Threads | Who's Online | Search

Posts: 99    Pages: 10    Prev 1 2 3 4 5 6 7 8 9 10 Next
Last Post
New Thread
This topic has been viewed 5832 times and has 98 replies
regallow
Re: Renaissance Technologies

Why wouldn’t you optimize an ETF model up until 2015 then use 2015 till now to get out-of-sample results. That way you would not have to wait several years to know if the model is good.
Valid point, Jim. I did not do that this time, but should have. I prefer to leave some recent time out of the modeling process until the end as a better approach than attempting to optimize factors, weights, and such, but sometimes I just don't.

Bob
"Logic is a systematic method of coming to the wrong conclusion with confidence." - Unknown

Dec 8, 2019 9:01:33 PM       
Jrinne
Re: Renaissance Technologies

This is what I found about Ray Dalio’s Pure Alpha Fund:

"Bridgewater’s Pure Alpha strategy has generated an annualized return of 12 percent since its inception in December 1991. But it posted a long streak of mediocre returns from 2012 through 2017, producing gains in the low-single-digits each year. "

RenTech RIEF in recent years:

"In 2013 it delivered 17.64%, in 2014 it brought back 14.53%, followed by 17.37% in 2015, 21.46% in 2016, 15.22% in 2017, and 8.5% in 2018."

I do not think the annualized returns of 12% and 11.5% (rounding error) for the Alpha Fund and RenTech RIEF can be compared due to different start dates. Perhaps they could be compared to each other if the excess returns were given (I do not have that data).

Recently anyway, RetTech wins in a fair comparison of long term investing strategies. At worst (for RenTech) one cannot prove one strategy is better than the other.

Clearly, RenTech has a viable long term quantitative investing strategy for its RIEF Fund.

It is possible that Ray Dalio is having some of the same problems we are when trying to compete with RT;-)

-Jim

From time to time you will encounter Luddites, who are beyond redemption.
--de Prado, Marcos López on the topic of machine learning for financial applications

Dec 9, 2019 5:01:25 AM       
Edit 4 times, last edit by Jrinne at Dec 9, 2019 5:13:52 AM
judgetrade
Re: Renaissance Technologies

Some stuff from my trading coach:

I think the first thing you need to do as a trader / Investor is to analyse your competetive advantage.
- Emotional Stability, Faith, GRIT
- No pricipal agent problem bc. you trade your own money
- Not to have to track the index
- relative small account size (size and liquidity premium is big!)
- P123, with a great cost/value structure, Great Community
- longer term horizon (not investing, not trading, something in between, a nice niche!)
- Beeing able to trade with minimal price impact (scaling in and out of position, Limit and GTC Orders, o.k. Broker)
- Knowleged about emotions of market participants and how to hack them with stable factors
- factor diversification

What I do not have:

- Besides PIT Data from P123 (which is a big comp. advantage!) I do not have (yet) Alternative Data (the biggest driver from R., their AI works super, bc. they have
tons of Alternative Data available before the others have it, that is 80% of their edge)

- Intersting way to fund my leverage (And I am glad, bc. big leverage is a high prob. death in the future, no matter how you structure
your trades! If your shorts go up and your longs go down and your are leveraged you are done, I believe that even R. underestimates
that kind of risk, a lot can happen in 100 Years.)

Its o.k. to look what R. or for example prop Trading firms do and get inspired. But at the end of the day you
got to trade what is in front of you (your existing completetive advantage!)

I know the grass is always greener on the other side, but from my experience this is dangerous thinking.
"Just that one improvement of my trading system and I will be fine. And then the next, and the next, and the next."
You will never find a system that is a straight line with a Sharpe of >4 and no DDs.
That will probably not be reached by a lot of us. Deal with that!

So what now?

Work on your mental state and be able to trade what is in front of you! I would be happy to cost average into the SPY and I am thrilled
on what modells I created on P123. Value that!

Try to get rich slow, not fast and teach yourself to be able to get through DDs and be thankfull on what you have and
economize as much as possible and invest it in your port.
That is 80% of the game (20% is the trading system!!!!!!!!!!!).

An investor with that mindset with a not so good strat will outperform somebody with a much
better strat. Its not what you know in investing, trading, its how you behave consistently over 25 Years that makes you successfull!

I know Prob traders that made > 100%, 500k a year. They blowed the money and did not reinvest it.
Their edge went away and they where out of business (I can related, I did something similar from 1996 - 2003).

If you are north 10-20% ann. all you need is time! Be thankfull and have faith, its important!
In the meantime you can work on your edge, but know your competetive advantages and sharpen them (the mental side as much
as the system).

Best Regards

Andreas

Dec 9, 2019 5:18:54 AM       
Edit 1 times, last edit by judgetrade at Dec 9, 2019 5:20:48 AM
Jrinne
Re: Renaissance Technologies

Some stuff from my trading coach:

Andreas


Me, if I were trading the black-box strategies (equal weight) from some of the Designers I would need more than a coach. Maybe some medication would help. But a coach would not cut it.

Maybe a therapist to help me understand why I would do such a thing in the first place would be the thing.

Everyone is all for screening stocks then understanding them fundamentally, Fine. But then P123 has a business model that advocates automated black-box strategies that have always performed below their benchmark, on average.

Maybe someone should just admit me—the world (or me) has gone mad. But until they shock me or give me a lobotomy I cannot see why we would not want to diversify this model.

We are all for Fundamental Analysis until it is time to invest in a black-box created by a retired Engineer or IT guy in a basement somewhere whom we have never met.

I do like P123 a lot—especially since I am that guy in the basement. That guy with no formal training in any pertinent field.

P123 is a great manually optimized quantitative investing tool. Committed to making not a single advancement in the quantitative strategies market for some reason.

-Jim

From time to time you will encounter Luddites, who are beyond redemption.
--de Prado, Marcos López on the topic of machine learning for financial applications

Dec 9, 2019 5:34:20 AM       
Edit 14 times, last edit by Jrinne at Dec 9, 2019 10:18:22 AM
geov
Re: Renaissance Technologies

Me, if I were trading the black-box strategies (equal weight) from some of the Designers I would need more than a coach. Maybe some medication would help. But a coach would not cut it.
-Jim

One has to agree with Jim.
I made the exercise to check the 5-year returns of the Designer Models.
There are 75 DMs with inception 5 years ago or earlier.

The average total return over 5 years of the 75 models= 23.4%.
and the excess 5-year return over SPY= -43.8%

There are only 6 models which outperformed SPY over 5 years.
For the 6 models the average excess 5-year return over SPY= 25.8%

That is not great. We have to do better than that.
Perhaps good market timing is the answer.

Excel file is attached.

Attachment P123_R2G_List 12-9-2019.xlsx (60095 bytes) (Download count: 19)


Dec 9, 2019 5:00:08 PM       
Doug
Re: Renaissance Technologies

There are only 6 models which outperformed SPY over 5 years.
For the 6 models the average excess 5-year return over SPY= 25.8%

That is not great. We have to do better than that.
Perhaps good market timing is the answer.


Hi Georg,
I am curious, why do you think we (the collective "we") can develop timing models that add excess returns if "we" can't do it with non-timing Designer Models? I thought it was widely accepted that timing the market is incredibly difficult to do.

Do you have a method or a model that has provided excess returns over a significant time frame, out-of-sample?

Thanks.
Doug

Dec 9, 2019 5:46:52 PM       
Edit 1 times, last edit by Doug at Dec 9, 2019 5:48:44 PM
Jrinne
Re: Renaissance Technologies

x

From time to time you will encounter Luddites, who are beyond redemption.
--de Prado, Marcos López on the topic of machine learning for financial applications

Dec 9, 2019 5:56:45 PM       
Edit 17 times, last edit by Jrinne at Dec 10, 2019 4:22:24 AM
geov
Re: Renaissance Technologies

I am curious, why do you think we (the collective "we") can develop timing models that add excess returns if "we" can't do it with non-timing Designer Models? I thought it was widely accepted that timing the market is incredibly difficult to do.

Do you have a method or a model that has provided excess returns over a significant time frame, out-of-sample?

Thanks.
Doug

The best I can offer with no market timing is the Best10(VDIGX)-Trader. It selects 10 stocks from Vanguard mutual fund VDIGX according to a fairly simple ranking system. This has been live since October 2014 reported on our website iMarketSignals. The universe is updated every 3 months with the holdings of VDIGX.

As one can see it its excess 5-yr return over SPY = 38%, with lower max D/D. It also had higher return than SPY each calendar year. There has been discussions of this in the forum before. It's not a secret - anybody can do it.

Attachment VDIGX 5-yr performance.png (146220 bytes) (Download count: 74)


Dec 9, 2019 6:46:13 PM       
Doug
Re: Renaissance Technologies

Thanks Georg,

I don't think my question was very clear, sorry. I was interested in a market-timing model that has provided excess returns over a significant time frame, out-of-sample. I would love to subscribe to that, or learn to build my own. Thanks.

Dec 9, 2019 6:54:28 PM       
geov
Re: Renaissance Technologies

I agree with both of you. I have started to look at timing—but not in or out of the market completely for me. Rather switching to different strategies or sectors e.g., XLP to XLK. I have no evidence that this a good strategy but I do not discount it. And the market is proving to be cyclical. I think none of this is new to Geov.
-Jim

Jim, QQQ-XLP works better.
Here is the backtest with my market timer formula $portsum.
I will explain the market timing strategy later.

Attachment aggregate timer QQQ-XLP.png (223972 bytes) (Download count: 68)


Dec 9, 2019 8:56:20 PM       
Edit 1 times, last edit by geov at Dec 9, 2019 8:58:07 PM
Posts: 99    Pages: 10    Prev 1 2 3 4 5 6 7 8 9 10 Next
 Last Post