Index | Recent Threads | Who's Online | Search

Posts: 12    Pages: 2    1 2 Next
Last Post
New Thread
This topic has been viewed 317 times and has 11 replies
sthorson
Expected Annualized Return

What is your expected annualized return from your model/simulation/back testing, say 5 years annualized?

Much effort is put into creating a profitable model, so what return are you hoping to achieve?

Is it a firm number, like 10% or 20% or 5% annualized?

Or are you simply looking to do better than a benchmark by some amount? If so, how much outperformance of the benchmark are you targeting? +5%, double?

And last, have you been successful at meeting your return expectation?

-Steve T

Sep 4, 2022 8:48:54 PM       
Edit 1 times, last edit by sthorson at Sep 4, 2022 8:49:38 PM
rtelford
Re: Expected Annualized Return

Steve,

Generally backtesting any model or strategy, I'm targeting 40-60% annualized. I deal mostly in small and microcaps, and models typically have a minimum of 20-25 stocks, decent liquidity, etc. I've learned the hard way about over-optimizing, so I generally try and manage the constraints.

For "all-season" strategies, I backtest from 1999 - if this performance particuarly persists in the last 5-10 years (pre-COVID), even better.

For more "seasonal" strategies (those tuned in to the current market conditions), backtesting over the full period since 1999 has limited use, but rather the key is testing in similar market conditions. This is again difficult as you have fewer periods to work with, reducing the statistical significance. That said, when they work, they really work.

Achieving the same backtest performance in realtime/out of sample can be difficult, if not impossible. Still, the higher performance, the better (assuming strategies have a reasonable # of holdings and liquidity as mentioned above).

In terms of realtime performance, I've been fully invested in the market since early 2020 (just before COVID) using the philosophy above. From 2016-2018 I was using overoptimized models, with very inconsistent performance. I regrouped in 2019 and relaunched in January 2020.

Annual returns in 2020 & 2021 were 50% and 58% respectively. YTD I am invested in 6 strategies, performance ranging from 2% to 55%. Come year-end I usually determine overall performance.

Cheers,
Ryan

Ryan Telford -- also find me at:
Seeking Alpha
Twitter

Sep 5, 2022 4:53:16 PM       
sthorson
Re: Expected Annualized Return

Ryan, Thanks for replying! Seems your regrouping worked out for you, well done.

Small and Micro were certainly the place to be in 2020 & 2021.

My results for Live Strategies:

2020: Micro +34%, Small +23%
2021: Micro +31, Small +40%

Can't complain.

I also run a Multi-Cap Strategy which pulled +44% in 2020 and +27% in 2021. All of my portfolios will hold 18-22 positions.

The point of my original post wasn't to boast return, but rather, what your "expected return" is when you build a model?

Sep 5, 2022 5:33:55 PM       
rtelford
Re: Expected Annualized Return

Steve, good performance!

Apologies if the above came off as a "boast", rather I was attempting to compare my targeted model/sim target to OOS. On expected return on models & sims (before investing), I mentioned above that I target 40-60% annualized, while trying to keep the model as robust as possible (which is an entire discussion in itself). For "all-season" strategies I try and strive for this over the full 1999-today period; if that is achievable particularly thru the last 5-10 years, then even better (and excluding the COVID period). My own experience, anything higher than that has resulted from over-optimization. Not a hard and fast rule, but I do limit my "out of sample" expectations beyond this target.

"Seasonal" strategies can vary, but I would say 50% annualized is the rough ballpark I'm looking for in the backtested sim performance, ideally if it's something that can be tested over multiple, but similar market conditions. For some of these types of strategies, actually very poor performance in the "off-season" can be a good sign as well, suggesting that the strategy has been truly seasonal. One example of this type of strategy is crisis investing, which focuses on small cap value during times of crisis. I wrote about this here.
https://seekingalpha.com/article/4339779-quan...-composite-ranking-system

Hope that answers your question, if not I can elaborate further.

Cheers,
Ryan

Ryan Telford -- also find me at:
Seeking Alpha
Twitter

Sep 6, 2022 4:22:51 AM       
carymac07
Re: Expected Annualized Return

Thanks to you both for inspiring me.

Cary

Sep 6, 2022 3:23:36 PM       
sthorson
Re: Expected Annualized Return

Let me rephrase the question...

While I imagine most of us have a desired return expectation they would like to hit, which is the crux of my original question, the most realistic expected return is some percent over benchmark.

What is your goal of expected return over benchmark?

I'm guessing most would simply like to perform better than benchmark. Investment managers want a return over benchmark including fees so they are adding value.

In other words, we can build screens, simulate, back test and rinse/repeat. What % number are you trying to hit that makes the work worthwhile?

Sep 6, 2022 8:09:19 PM       
RTNL
Re: Expected Annualized Return

Ryan, Thanks for replying! Seems your regrouping worked out for you, well done.

Small and Micro were certainly the place to be in 2020 & 2021.

My results for Live Strategies:

2020: Micro +34%, Small +23%
2021: Micro +31, Small +40%

Can't complain.

I also run a Multi-Cap Strategy which pulled +44% in 2020 and +27% in 2021. All of my portfolios will hold 18-22 positions.

The point of my original post wasn't to boast return, but rather, what your "expected return" is when you build a model?


Steve and Ryan - that is fantastic! Can you shed some light on:
- total number of holdings
- Turnover and
- How has this year worked out so far?

RT

Sep 6, 2022 8:15:28 PM       
yuvaltaylor
Re: Expected Annualized Return

Let me rephrase the question...

While I imagine most of us have a desired return expectation they would like to hit, which is the crux of my original question, the most realistic expected return is some percent over benchmark.

What is your goal of expected return over benchmark?

I'm guessing most would simply like to perform better than benchmark. Investment managers want a return over benchmark including fees so they are adding value.

In other words, we can build screens, simulate, back test and rinse/repeat. What % number are you trying to hit that makes the work worthwhile?

I think you're right. If we can't beat the benchmark, then we might as well just invest in SPY. There's no point in putting all the work into this if we can't consistently beat the market. Personally, that's my goal.

There's also the question of past performance. That sets a bar for me. I want my current performance to be as good as my past performance. That's a high bar since factors get arbitraged and transaction costs go up with larger AUM and you can't put 15% of your money in a nanocap if you're managing millions.

So I keep a little chart that measures my six-month excess return (over a benchmark consisting of the stocks in my universe equally weighted) every month. If it generally trends upwards and rarely goes below 0%, then I'm very happy. (At the moment it hasn't dipped below 15% since the pandemic bear market. I had a horrible 2019, but everything's been coming up roses since. Prior to the pandemic, it was above 15% less than half the time.)

Attachment 6-mo excess return.png (20213 bytes) (Download count: 41)


Yuval Taylor
Product Manager, Portfolio123
invest(igations)
Any opinions or recommendations in this message are not opinions or recommendations of Portfolio123 Securities LLC.

Sep 7, 2022 10:02:23 AM       
Edit 1 times, last edit by yuvaltaylor at Sep 7, 2022 10:05:28 AM
sthorson
Re: Expected Annualized Return

Yuval, that's a great visual, and also very good performance.

In looking at your 2019 period I admit I struggled that year as well. I'm still trying to figure out why, so its good to see I'm not alone.

Sep 7, 2022 10:51:12 AM       
rtelford
Re: Expected Annualized Return


Steve and Ryan - that is fantastic! Can you shed some light on:
- total number of holdings
- Turnover and
- How has this year worked out so far?

RT


Most of my strategies have typically 15-25 stocks. I started out using 5 stocks portfolios, experienced some early success, only to find very poor out of sample performance shortly thereafter. 5 stock sims provide limited # of transactions to test your ideas.

Most of my strategies are rebalanced weekly, but some weeks go by without any turnover. Generally annual turnover in the neighbourhood of 150-300%. The more weight to price-related factors (P/E, P/S, etc), the higher the turnover.

This year has been rough on balance. My strategies were buying energy stocks at the beginning of the year, caught the surge in energy prices, then got hit by the drop. They've slowly been selling off energy holdings. My two "all-season" microcap strategies have struggled as of late. Despite being highly ranked, some very high ranking stocks have been hit very hard.

My performance is all over the map, of my 6 live strategies, YTD performance ranges from -0.5% YTD to 50%.

Ryan Telford -- also find me at:
Seeking Alpha
Twitter

Sep 8, 2022 4:53:22 PM       
Posts: 12    Pages: 2    1 2 Next
 Last Post