Thank you very much for sharing your ETF model simulation screen at P123, It is inspiring very much. We wish you to become more successful.
I need to work on my ranking to reduce draw down.
It is yearly rebalance, i have more room for improvement. most of the investment book preaches, yearly or long term investment, i can get some more ideas to work on.
Here, performance of my sim, only one sell rule, rank based.
Hemmerling,
Nice ideas, inovestor and you, both of you only accepting you are working for investment firm, sharing your methodologies and success periodically.
believe, you people only can promote p123 models and designer skills to get opportunity from open market.
=================================
agree DM market space is very small, only few 10s of paid subscribers.
=============
Kumar
Do you have any more models that might be appropriate for institutional investors?
Miro,
as Hemmerling said, the model take more time develop, if we have real investor requirements and his interest to buy models; we can work in that direction.
As an individual investor/trader, i would like to have 10 stocks models and weekly rebalance, 3X turnover is not big concern as we have opportunity to get 2X return of
institutional investors models.
Do you work with investment firm and managing others money ?
Inovestor brings up several good points, some that I have posted about before. The two biggest issues are the market cap and the hold periods. All the DMs are pretty much weekly because they tend to have the highest backtested returns and that is what people are trying to sell. Weekly trading is not practical for some, and impossible for others like myself. I have to hold stocks for 30 days minimum. Even if I could trade more frequently, moving a few hundred grand in and out of 10-20 nanocaps every week is a disaster. I tried it for a while chasing the 50-60% returns I thought I could get. They were fantasy.
I am much wiser now, well I like to think so. I put the bulk of my dollars into R1000 or S&P stocks. I have 10 screens that get me 20-25% annualized on a monthly rotation. Sometimes this is still tough to keep up with. I only follow Cherrypicking the BlueChips at this point as everything else doesnât work for me.
I think P123 should have a feature where people bid on stuff like this to encourage people to develop models based on real demand. Like eBay for PM. Then I think you would get some serious subscribers, maybe even institutions of hedge funds. I would pay a pretty penny for someone to get me a screen that starts with the R1000, rebalances quarterly, holds at least 20 stocks and gets 25+%. I think it is achievable, just not by me so far. As an example, say I bid $100 a month for this model and then you find 9 others with similar views and you would make $12k for the effort. Only a handful of DMs are even full because most donât meet anyoneâs specific needs.
The rub here is the return; everything else is an input constraint. Since simulations are notoriously easy to exaggerate, how much out-of-sample data would you need to see before deciding that the return design goal was met? Letâs face it, selling the sizzle and not the steak works and is still a common practice on P123.
I love the idea of surveying the input needs of potential DM subscribers, but would their return goals be realistic? I mean, given the S&P 500 as a working universe and its 9% annualized return, where would prospective subscribers call BS? 15%, 20%, 30%, 40% AR? Personally, less than 20% sounds realistic. I would love to find a DM offering more, but it would need to demonstrate that with 6-12 months of out-of-sample data.
Miro sent me an email this morning about joining this group since Iâm an RIA and do manage institutional portfolios and have come very close to what others have been asking in this forum for mid to large cap growth portfolios with a 25% return. My 5 year return on Collective2.com for a 20+ holding portfolio is 22% gross and 20.5% net.
It seems to me there is always needs to be a realism check with P123 subscribers. First, 25% CAGR is difficult but not impossible but it depends on the market environment. For a model to produce 25% returns youâd need a major long term bull market like we had in the 80âs and 90âs. Fidelityâs Magellan did approx 18% per annum in the 80âs. If a secular bear market occurs you wonât make 20%+ percent.
In addition, the returns will require stocks that have good volatility and or beta and big winners. Youâll need to dispense with the reliance on Value stocks and look to Growth/Momentum.
Secondly, someone said âtheyâd pay a pretty pennyâ maybe so but a portfolio with such Alpha will be charging a pretty penny and I charge 1.5% per annum (under $500K) or its available on C2 for the same rate. I adjust the fee dependent on the assets linked.
Lastly, I left P123 because I think most developers here are just kidding themselves to think that rigid fixed programs can deliver that kind of return. When I mean âkidding themselvesâ what I especially mean is that youâre trying to develop systems that Goldman or every other prop trading hedge fund hasnât thought of or squeezed the Alpha from.
But I run an algo on the universe of stocks I look at but I pay very close attention to moving averages and volume. There is subjectivity in my management but its based on 30+ years of trading and it works far better than anything I ever saw or developed on P123.
Nice to see so much institutionals and pros here.
I am still happy that I can surfe below the radar of the institutional investors, since one of the best competetive advantages
is a small private portfolio. I hope it does not stay that way so hoping to get size problems one day tooâŚ
Well, Iâve decided to jump into the fray and launch an institutional-grade designer model with ultra-low turnover (55%), average holding time of two years, high liquidity (minimum median daily dollar volume of $3 million), no market timing, and 25 stocks. Its performance isnât as good as Kumarâs model, but I guess weâll see which one performs better out of sample in a few yearsâ time. This model prefers overlooked, undervalued, and stable companies with lots of cash, strong sales, negative accruals, and good growth potential; Kumarâs favors large-caps and mine favors small and mid-caps. You can see it here:
My two cents is that this is not a realistic institutional grade product (quarterly re-balancing, large-cap stocks, long-term 25% CAGR, no market-timing). I think this would be a better description of private money who wants a DM to have high liquidity. With all the constraints on an institutional model - I donât see how this could be developed without extreme curve-fitting on a small portfolio of 20 stocks or less. Or the 25% CAGR comes at the expense of extreme drawdown risk such as certain value situations. Either way, I donât see this as a viable institutional grade product. If it was, it would be worth far more than $12,000 per year. I would say 10 - 100x that much.
Yesterday I failed to mention that I have two more live portfolios that have stats requested in this forum (my portfolios are not hypo-backtests but real time model portfolios).
OP II (up to 20 mid and large cap stocks) has a 24.8% NET return average and 27.9% gross CAGR for the past 5 years.
Like the Vegan Growth Portfolio both have the same ethical screening criteria and its a Momentum strategy using Mid and Large Caps with the occasional high volume small cap. I like to trim losers very quickly and winners can stay in the port for as long as they continue to perform.
None of my portfolios including our OP I (up to 10 stocks) (which has a 5 year CAGR in real time of 35.4% NET and 38.8% gross CAGR use any sort of value oriented ranking system. I donât use any sort of designated rebalance period, stop losses, price targets,value metrics or fixed position sizing (which is lazy IMO)
All Iâm trying to say is that the expressed performance metrics in this forum are possible in real time. But most of the years I spent on P123 were not wasted as I learned a great deal. One major lesson was that heavy backtestins is your enemy and not your friend. I read someplace that complex systems could see a 75% drop in Sharpe ratios in real time and simple backtested systems could expect a 30% drop. was surprised at the percentage that failed.
It is my perspective
There are many mutual funds, ETFs and active managers with more funds for Institutional, and they have many rules to follow.
one of them donât use market timing.
Just beat the market,
even loss also acceptable if it is less than bench mark loss.
very low turn over, 2 years avg holding, look like buy and hold approach and passive investing.
as rebalance is yearly or quarterly, P123 weekly upto date valuable data become useless. the power of now is missing.
they have 15% to 17% return as successful model presentation in out of sample.
they have pressure to perform in that range 100% every year, to continue to attract clients with the same presentation.
As an individual investor, i can design model, what i wanted as consumer to my personnel account.
Fundamental and Technical both has negative and in extreme downtrend, i will be cash 100%.
10 liquid SP500 stocks is good to hold and manage for personnel account.
would like to get positive return every year and better than bench mark return every year.
3X to 4X turn over is at the average 3 to 4 months holdings is acceptable. between active and passive investing.
As rebalance is weekly, P123 weekly upto date valuable data is useful.
20% fees to P123 is justifiable.
My model in simulation shows 30% to 40%, In out of sample, 50% of the return simulation = 15% to 20% avg per annum also will be acceptable as successful DM relative with other liquid model in P123.
There is lot of chances to perform 25% to 30%/avg for 5 years period.
I can watch my model weekly basis closely to grow further in knowledge and designing skills.
I would like to continue to learn and grow as good designer at DM market.
I am happy, i am a better designer than last year and making a good progress when i find time.
I have more well wishers at P123 than last year.
I will continue to grow in DM space for individual investor with liquid 10 stocks models with reasonable turnover.
I am requesting Investment Advisor / Institutional Investment Professional to provide
some models in DMs for individual investor.
That way we will have more valid ideas, theme and real growth in correct direction.
Right now, Mgersteinâs Cherry Pick model is professional grade model backed by 3 to 4 decades of fulltime investment experience and few more good designers are in P123,
if Inovestor and many others IRAs can provide similar models with their own theme, P123 can prosper in coming years.
SP500 universe - 25 stocks with yearly rebalance return around 20% avg annual with 55% draw down - without market timing screens are posted a week ago in this thread.
the same ranking used for this below 10 stocks VGQ designer model, which will be available in DM by next week.
a. Weekly rebalance
b. Additional buy rules
c. Market timing used to go 100% CASH when SP500 death cross SMA(50) < SMA(200) and EPS in confirmed down trend.
no other economy indicator used for market timing as unemployment rate and interest rates are depends on FED policy & decision.
d. avg holding period 4 months. 30 transactions per year at the average.
NOTE: For novice investor, i am asking to wait for few more months out of samples before subscribing.
===============================================================
DISCLAIMER:
By subscribing to this Smart Alpha strategy you accept all terms and conditions set forth below, and confirm that you take full responsibility for your own trading or investment practices and decisions including which stocks to buy or sell and when. This Smart Alpha strategy is not making specific trading recommendations and should be viewed as information only that any investor may or may not choose to act on. Accordingly, you agree that you will not hold either this Smart Alpha developer or P123, responsible for any losses you may incur.
Perhaps worth noting than 25% realized annual returns over ten years would place a model among the top 1% of all mutual fund and ETF performance records over the last ten years, a period of near-historic market performance.
Start Your own Hedge Fund. This model works fully funded out-of-sample with 15 stocks and the results are statistically significant by any statistical measure you might chose. I have considered starting a 100-stock designer model with the name Start Your own Hedge Fund. ,
I think hedging the drawdown would be the main problem. It could be combined with some SP500 models to maximize liquidity. But you would have to use options or something to keep people invested during the drawdowns.
In any case, I think the number or stocks would not have to be an issue. I am not sure what to think about the turnover but not huge, perhaps:
