Backtesting based on fundamental data

[DeepState]

That's because you obviously haven't read Seth Klarmans "Margin of Safety" wherein he explains all of this at the beginning of the book. Some excerpts

re smallcaps


But more importantly on the general "who is the patsy at the table that craft takes money off when he's winning":



Klarman posits that the largest money in the market - instos - are indexing. This is literally antithetical to the idea of a stockmarket, i.e. capital formation for the companies with the highest ROC, where instead you invest in a company regardless of its ROC, or any fundamental factor whatsoever. In the index case whoever has the largest market cap (shares on issue * current price) is invested in most heavily. The more people who index, the less efficient the market becomes - stocks outside the index are underpriced and undertraded relative to those in it. You are often allocating more money to companies that destroy value than those who create it.

Not just that but there are plenty of other irrational, similar, reasons covered at the beginning of the book that explains where the largest market inefficiencies that a value investor can capture come from:

* Securities in the index are often playthings for speculators and wannabe arbitrageurs, resulting in significant overtrading relative to their fundamental value and this overtrading (while giving the appearance of liquidity) is often the source of return drag. Low liquidity stocks outperform high liquidity stocks across all market cap quantiles. Low liquidity value outperforms low liquidity growth, low liquidity smallcap value outperforms low liquidity large growth, and so on.
* Funds don't necessarily choose when they sell - long only funds often prefer to retain earnings than sell a security looking for another representing better value and capital formation properties so they will still be holding stocks at "the top" when Klarman is in cash. Leveraged funds might sell an entire holding on a drawdown that's 1 tick beyond their pain point. and so on....
I suggest reading the book.

Ha ha. You probably read at 2000 words per minute Sinner! Given the book has 76,736 words in it, you'd eat it in less than 45 mins! Crikey dude...please let me know the name of your speed reading coach. I am awed by your depth of knowledge and I used to read over 100 books a year. Obviously on stupid stuff. Sigh.

You're right. I have not read Klarman's book but think the way he and, thus, Baupost, invests are very interesting. His stock picking ability is out of this world.

Everything you have said is true. I agree with all of it. In terms of the efficiency of index vs active, I would add that there is the Grossman-Stiglitz 2003 "On the impossibility of informationally efficient markets" which also talks about the fact that alpha will be available for harvesting. It's just a matter of who is getting it. It remains available for insto despite his criticisms of how they might, in aggregate, seek to obtain it. He is highlighting peer/career/business risk concerns and these do lead to additional frictions - no argument there.

So, these results stand in the US and just about everywhere else. But they do not hold in Australia? Well, the small cap value premium exists. But I'm just talking about transfer of wealth defined as pure alpha - which is the thing that Baupost/Klarman is talking about. The market is sub zero-sum in terms of transfers. So that's the puzzling part. I agree with Baupost and what you are saying. But if it is true, why then are Aust insto doing so well vs index across the spectrum of cap? I don't get it. This result does not line up with Baupost/Klarman's thesis which I also subscribe to. That's why I am so puzzled. It is a puzzle that no-one that I know of has explained well although some postulates have been put up. In large cap, examination of tick level data shows that the patsy is foreign insto. Retail does alright. They get routinely wrong footed. But why do they stay so stupid (partly it is the thesis above, but also the kinds of stocks they hold and the way they mis-time their market movements)? But as you go down the cap spectrum, this also ceases to become a viable response because the ownership habitat changes - if insto is doing very well on average - who is the patsy in this market? Because it isn't insto. It's foreign insto in large cap, but it is not likely to be that as we move to smalls and micro - they are not active there. However, since retail does alright in large (for reasons other than alpha directional plays, I would add) I am not currently prepared to say it is retail and think I am missing something. But all roads presently lead there by tautology.

Hence the question. I don't know the answer and it seems to defy the postulates you have just added - which I agree with also. Something is exceptional about the Australian market that seems to favour domestic insto. So if the theory doesn't hold against the outcome...we need to look some more. And these results have held for decades.

Do you have a view?

 

[sinner]

So, these results stand in the US and just about everywhere else. But they do not hold in Australia? Well, the small cap value premium exists. But I'm just talking about transfer of wealth defined as pure alpha - which is the thing that Baupost/Klarman is talking about. The market is sub zero-sum in terms of transfers. So that's the puzzling part. I agree with Baupost and what you are saying. But if it is true, why then are Aust insto doing so well vs index across the spectrum of cap? I don't get it. This result does not line up with Baupost/Klarman's thesis which I also subscribe to. That's why I am so puzzled. It is a puzzle that no-one that I know of has explained well although some postulates have been put up. In large cap, examination of tick level data shows that the patsy is foreign insto. Retail does alright. They get routinely wrong footed. But why do they stay so stupid (partly it is the thesis above, but also the kinds of stocks they hold and the way they mis-time their market movements)? But as you go down the cap spectrum, this also ceases to become a viable response because the ownership habitat changes - if insto is doing very well on average - who is the patsy in this market? Because it isn't insto. It's foreign insto in large cap, but it is not likely to be that as we move to smalls and micro - they are not active there. However, since retail does alright in large (for reasons other than alpha directional plays, I would add) I am not currently prepared to say it is retail and think I am missing something. But all roads presently lead there by tautology.

Hence the question. I don't know the answer and it seems to defy the postulates you have just added - which I agree with also. Something is exceptional about the Australian market that seems to favour domestic insto. So if the theory doesn't hold against the outcome...we need to look some more. And these results have held for decades.

I am not sure I understand what you're saying here?

There is this question:

why then are Aust insto doing so well vs index across the spectrum of cap?

Given the appropriate data you should be able to answer this question with a Principal Component Analysis, i.e. define a bunch of fundamental factor return paths and find which (or combination of which) have the explanatory power against the insto return paths.

But I would ask:

* Which index is the benchmark (XAO or XJO)?
* Which instos do you include and exclude from this statement?
* Are they actually outpeforming on a Sharpe/Sortino ratio basis?

But also you seem to be asking another question about why certain market cap quantiles might outperform others?

But as you go down the cap spectrum, this also ceases to become a viable response because the ownership habitat changes - if insto is doing very well on average - who is the patsy in this market? Because it isn't insto. It's foreign insto in large cap, but it is not likely to be that as we move to smalls and micro - they are not active there

Can you clarify this?

To me the answer to the second one is simply an issue of liquidity, i.e. overtrading.

 

[craft]

I am not sure I understand what you're saying here?

Me either, but I'll give you a view on what I think you are asking.

But I'm just talking about transfer of wealth defined as pure alpha - which is the thing that Baupost/Klarman is talking about. The market is sub zero-sum in terms of transfers.

That is only true of the entire market – it does not have to be for the sub index’s

So just restating what McLovin eluded too because I think he is right.

The top 100 stocks in the Small ordinaries are simultaneously the bottom 100 stocks in the ASX200.

Small funds measuring themselves against the small ordinaries don’t have to be zero sum (gross of fees) in total for outperformance against the XSO. Their overall outperformance/alpha can be soaked up by funds measured against the XJO. Given the difference in amounts benchmarked against each index a reasonable overall outperformance by small cap managers may only register as a small underperformance to the funds (possibly foreign) bench marked against the XJO.

 

[DeepState]

1. I am not sure I understand what you're saying here?


2. Given the appropriate data you should be able to answer this question with a Principal Component Analysis, i.e. define a bunch of fundamental factor return paths and find which (or combination of which) have the explanatory power against the insto return paths.

3. But I would ask:

* Which index is the benchmark (XAO or XJO)?
* Which instos do you include and exclude from this statement?
* Are they actually outpeforming on a Sharpe/Sortino ratio basis?

4. But also you seem to be asking another question about why certain market cap quantiles might outperform others?

Can you clarify this?


5. To me the answer to the second one is simply an issue of liquidity, i.e. overtrading.

Great questions.

1. In am trying to understand why, in Australia, insto outperforms the cap weighted index for accounts managed against ASX 200/300, ASX Small Ords and even for indices in Micro cap. It implies that they are in receipt of transfers of alpha from others in the market, in aggregate. This has been going on for decades and defies what Baupost, you, Craft, McLovin and I think should happen.

2. PCA is used for explaining cross-sectional returns and the factors which emerge are blind although you might be able to template them against fundamental factor return time series. So the argument goes that insto may be holding a misweight to certain factors relative to index. These factors must show non-zero/positive mean. It may be possible to find out what these factors pertain to even through this will always be approximate. Absolutely.

Sure. I guess part of it may be due to permanent factor bias in the insto sample. But, if you examine the universe, you have pretty much all the major styles accounted for (value/growth/quality/mom/cap/leverage/liquidity...). In aggregate, these exposures are close to flat. What is left, if you observe the smalls outperformance, is a really big excess return that cannot possibly fall within the ambit of a factor bias. It could be factor timing, but that's alpha. It isn't due to permanent bias. Well, some of it might be, but no way to that extent. In the large caps, which are benched to ASX 300 and ASX 200, the insto manager universe is massive. There are nearly 100 managers surveyed and anyone with $20m or more who wants to show numbers in order to get and retain clientele is in there. Selection bias has been shown to be very small. Difference between equal weight and cap weight also is not a big thing. The median manager routinely produces returns above the indices by ~2% per annum. ie. the major factors are flat-ish, yet you have outperformance of index.

Hence, it largely falls to stock selection. The bit not explained by key fundamental factors, or not explained by beta (if only using PCA for average manager vs index) or the higher order Eigenvalue PCA components if looking at the cross section of the index and the manager universe - although this is hard because the universe changes. So it's still weird.

3. Great question. I do not know. But apart from Low Vol and a few exceptions like this, most of these guys are seeking to outperform the index in absolute terms. They have risk stats all over the place. Aggregate beta is generally close to 1 although this varies a lot with some at 0.7 ex post, to others in the 1.1 range. Risk adjusted or not, insto is taking money out of the market. I guess you are arguing a segmentation approach may exist...insto seeks one type of risk-reward outcome and others in the market seek another and all can be happy because it's not win-lose but win-win. That's certainly a viable explanation, but why is Australia out of whack with the rest of world? Presumably the rest of world has the same segmentation arguments as well and their mix of objectives should not (I guess) be that different to what exists in Aust. Aust managers are seeking to beat the index and, typically, each other. Bottom line is that they are taking money out of the market vs others. In general these others presumably don't intend to lose money vs rest of market on average. Yet they are....in Australia.

4. Craft has argued that this may be due to a capitalisation banding issue. McLovin has indicated that his edge lies in his participation in stocks below the radar on the basis of capitalisation.

Craft's argument says that the Small Cap managers might be outperforming. Because of index overlap with large cap in the ASX 100-200, it is possible that the Small Caps outperform and take alpha off the Large caps and that they can also outperform. However, this still results in the outcome that insto is outperforming in aggregate vs rest of market even if Small cap insto is taking money off large cap insto in this example (and the obverse of large cap taking money off small cap and both outperforming can also exist). Either way you take this argument, insto is taking money over others. This extends to the micro caps as well. A range that extends below the level of even McLovin's search zone. Hence, we really can't argue along the lines that it is below the radar. At every capitalisation band, insto is taking money off the rest of the market.

5. Liquidity demand and t-cost may be a reason why others add to the zero sum nature of aggregate performance. But frictions of this nature don't impact the market. The index return does not factor in any t-costs. Are you arguing that insto is providing liquidity to the rest of the market? That would be a viable explanation, perhaps. However, it has been found that insto generally demands liquidity. It turns out that retail is the liquidity provider and gains from this. Pretty amazing. So liquidity provision is probably not a big explanatory factor for the level of outperformance seen. The active exposure to the liquidity factor is not significant enough, nor the returns to the factor high enough, to make up for the difference in any way.

So I'm still amazed and puzzled. My best guess is that insto gets preferential treatment for seasoned offerings in bookbuilds which don't exceed about 10% of the existing float - depending on company rules and ASX rules. Because of index construction rules, they are usually not added in index weight until much later. The insto managers get to pick up the premium and the index doesn't show anything. Insto has this access and it is not available to retail. But but but...this stuff must happen overseas too. Puzzled again.

 

[craft]

The chart you referenced below was for small cap manager’s excess return.

Are you now saying that all managers in aggregate produce an excess return over the total market return?

What’s the supporting evidence for that claim?

Private placements (Grrrrrr) – I’m not sure how its extent of use compares to other markets but they seem to get used and abused here relentlessly. Maybe part of the answer????

 

[DeepState]

The chart you referenced below was for small cap manager’s excess return.

Are you now saying that all managers in aggregate produce an excess return over the total market return?

What’s the supporting evidence for that claim?

Private placements (Grrrrrr) – I’m not sure how its extent of use compares to other markets but they seem to get used and abused here relentlessly. Maybe part of the answer????

Hi Craft

Yeah, right across the spectrum. Maybe it wasn't clear enough in Post 3 May, 11:10pm.

Yep, there are surveys around from, say, Mercer, which have recorded sector returns since the mid 1990s. I don't have one to hand but they have shown consistent outperformance of the broad market index since their inception - for Australian equities. The rest of the sectors are much more in line with what we'd think should happen. Mercer is the "Go-to" place for such things. If you aren't on the database, like the tree falling in the forest, you never really happened. So the participants in the survey essentially contains any insto seeking external assets.

I think SEOs are part of the answer. They just aren't all of it, or even the majority of it. The volume offered and discounts offered just aren't large enough to make a decent impact on the gap we are looking at. It's a pure benefit to instos for just turning up...and getting lunched.

Cheers

 

[DeepState]

Managed to drag one out. Australian equity survey from Sept 2013. Please note that the number of managers surveyed exceeds 100. Although the surveys are point in time, where it might be considered that survivorship bias exists (see how the number of managers increases as the time period shortens), this has been investigated and not found to be material in skewing results. The figures (median - index) to 2013 are weak relative to history but still show material outperformance in aggregate. This does not occur elsewhere in major markets - to my knowledge - for reasons as previously posted. Nonetheless, here they are. The columns are 3mth, 1yr, 2yr, 3yr and 5yr ended 30 Sept 2013 annualised for periods longer than a year:



Source: Mercer

Mercer takes other actions like limiting the length of historical data accepted into the survey to a very short period etc. to control for look back bias. As managers move to extinction, they still tend to report right to the end in the hope of getting money and, whatever happens, the record is still preserved...etc..

 

[sinner]

Managed to drag one out. Australian equity survey from Sept 2013. Please note that the number of managers surveyed exceeds 100. Although the surveys are point in time, where it might be considered that survivorship bias exists (see how the number of managers increases as the time period shortens), this has been investigated and not found to be material in skewing results. The figures (median - index) to 2013 are weak relative to history but still show material outperformance in aggregate. This does not occur elsewhere in major markets - to my knowledge - for reasons as previously posted. Nonetheless, here they are. The columns are 3mth, 1yr, 2yr, 3yr and 5yr ended 30 Sept 2013 annualised for periods longer than a year:

View attachment 57871

Source: Mercer

Mercer takes other actions like limiting the length of historical data accepted into the survey to a very short period etc. to control for look back bias. As managers move to extinction, they still tend to report right to the end in the hope of getting money and, whatever happens, the record is still preserved...etc..

Hm. That seems quite odd, if I'm understanding the table correctly, to say the least. Not to mention, completely at odds with not just the global contemporary situation as you mention, but even my own understanding of the Australian market?

I present the following paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2339661

We investigate the existence and sources of performance persistence among Australian equity funds, using monthly portfolio holdings data. We find that persistence exists, and that it relates to outperforming rather than underperforming funds, primarily stems from security selection skill, and is associated with exposure to growth and high momentum stocks. Further, persistence largely derives from existing holdings, while subsequent active trading contributes only moderately positive returns for both outperforming and underperforming funds. We also find that persistence fades beyond six-months after the initial ranking, and vanishes after 24-months. In summary, persistence amongst our sample largely stems from the medium-term security selection and portfolio construction skills of previously outperforming growth-orientated managers. Differences between these findings and those for U.S. equity funds imply that the existence and nature of persistence may depend on market context.

(the sections highlighted in red conform with my understanding)

and also "The Case for Indexing Australia" - which I guess should be taken with a grain of salt, since it's a Vanguard funded study and of course they'd be pro indexing, however the data is from Morningstar and I have no reason to assume they've fudged these results:

https://static.vgcontent.info/crp/i...se-for-Indexing-Australia.pdf?20140324|093000
The entire link is worth reading to discuss the differences between your understanding, and the one presented within, I also provide one of the many useful charts contained within

 

[DeepState]

Hm. That seems quite odd, if I'm understanding the table correctly, to say the least. Not to mention, completely at odds with not just the global contemporary situation as you mention, but even my own understanding of the Australian market?

I present the following paper:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2339661

(the sections highlighted in red conform with my understanding)

and also "The Case for Indexing Australia" - which I guess should be taken with a grain of salt, since it's a Vanguard funded study and of course they'd be pro indexing, however the data is from Morningstar and I have no reason to assume they've fudged these results:

https://static.vgcontent.info/crp/i...se-for-Indexing-Australia.pdf?20140324|093000
The entire link is worth reading to discuss the differences between your understanding, and the one presented within, I also provide one of the many useful charts contained within

View attachment 57873

Hey hey...

Gee you dig well.

On the SSRN paper...I actually know those guys. One used to broker to me and chat weekly even after I retired, another bought our stuff, another is one hell of an academic another is also a darn fine academic and I don't know the person who probably did all the work (the last name). It's an all-star ensemble.

The research there is about relative performance persistence. They find that there is persistence on the most successful funds. Other studies conducted elsewhere show persistence on the underperforming funds (because we can figure out what people are holding over time and kill them. It's also for reasons of what fund managers do when under intense selling pressure). Their findings are perfectly fine by me. There is momentum in the Australian market. It is the strongest factor. Although it is much weaker elsewhere and negative in Japan.

They show autocorrelation in the cross section. It does not explain why the mean is above zero. It explains why funds rotate around the mean.

So the conundrum continues.

I also looked at the Vanguard figures. The Morningstar data is net of fees. See the footnote (2) on page 4. Whilst relevant, the question I have relates to the pure alpha extraction. Before fees but net of frictions, domestic insto are extracting alpha vs rest of world.

 

[sinner]

Glad you appreciate the digging

Last one from me, it's a stumper and I'm just not qualified to answer

http://afrsmartinvestor.com.au/p/shares/what_fund_managers_won_tell_you_8MYjF4vobqdBitF1BSocaP

The message of this article is that fund managers underperform. It notes

Countless studies have cast doubt on the benefits of active management in highly efficient markets, especially US equities, where competition is high and (in theory) prices immediately reflect all publicly available information.

While the Australian equity market appears similar, the evidence is less compelling but still gives much cause for scepticism.

(personally my opinion is the difference in how compelling the evidence is down mostly to lack of research and available sample size as the market is that much smaller/younger)

but later in the article a snippet from Mercer

Australian Super’s chief executive officer Mark Delaney told Smart Investor there were maybe “two or three” active fund managers in *Australia who can consistently beat the index by three or four percentage points.

Simon Eagleton, a senior partner at Mercer, argues there are a number of factors that allow funds management professionals in the Australian equities market to outperform compared to the US market.

For example, the Australian market has a greater number of less know*ledgeable retail investors trading, as well as foreign investors chasing broad thematic investment strategies, such as resources, which may not be connected to fundamental stock prices.

“That can create an inefficiency that perhaps local professional investors can exploit,” he says.

I can definitely see an interesting point there about the broad thematic investment strategies from overseas, perhaps foreign is more active in the space (via a less official/trackable investment vector?) than you thought?

 

[DeepState]

I can definitely see an interesting point there about the broad thematic investment strategies from overseas, perhaps foreign is more active in the space (via a less official/trackable investment vector?) than you thought?

Delaney is referring to a very high hurdle. I'm not going to argue that one. He's entitled to his view.

Simon's reference to dumb retail is actually not correct. He does not have access to the data to discern what is going on there. I thought so too until we did the work and found otherwise. It's an easy assumption to make.

His references to foreign insto playing themes is correct. They see Australia as a bunch of banks and resource companies. Foreign insto invests in these thematically for the most part. When they move in, though, they stick to the major names for the most part. Just an anecdote but if you look at the ownership of NCM-AU it is loaded with offshore insto. Look at KCN-AU and it is dominated by local insto. Same theme, vastly different ownership profile. Same sort of thing, but less strong, for WBC-AU vs BEN-AU.

So foreign insto as patsy in large cap is a contributor. Why it keeps happening is a mystery to me. It is a less viable explanation when you move down the cap spectrum. So I still don't know who the pasty is there. If this goes any longer, it's obvious it is me....

 

[KnowThePast]

Sorry for the delayed reply,

Back in the day, we had to build our own database because commercial systems miss a bunch of things and have problems with primary keys. Stock codes would be re-used etc. We also had to classify stocks into the correct sectors too.

Fundamental data for 3-way accounts was obtained from a range of data vendors in historical terms. These had to be compared and corrected. Estimates data was obtained from the brokers who were producing those estimates. Proprietary data was sourced from wherever it needed to be sourced from.

The backtesting engine was built in MatLab. The database was SQL and visible through PowerBuilder as well as via SQL query. C# was in there as well. MatLab was also used to form portfolios by manipulating all this data into a trade list. Optimisers in MatLab, though awesome, were not awesome enough for these purpose. Hence optimisation was linked to Lindo Pro which provided commercial grade stuff useful for backtesting and real portfolio construction.

All proposed trades were run through a compliance check via extraction of holdings via PowerBuilder from HiPortfolio files. These were compared against client restrictions. The HiPortfolio files were also source of truth for portfolio construction with information from back office relating to overnight cash received or intra day cash receipt. We also loaded information parcels where a client wanted tax aware portfolio management which were coded for tax rate and accounting convention applicable. These were considered in the trade off between forecasted returns and tax costs amongst other things.

Trade feeds via DFS IRESS allowed direct communication of trades and bookout without manual entry and intra day management of the trading activity of the day.

Then we sipped Pina Coladas and had long lunches. Or, sometimes, took a dive off the Barrier Reef or Bermuda.

I had a very busy summer here, neglected my favourite topic on the forum. Some very interesting discussion here and I wanted to bring it back closer to the thread title.

RY, I love hearing everything you have to say, any other backtester developers on here, I'd love to hear from you.

As an amateur, I wish I had the means to pay for data, it vastly increases the scope of what is possible to backtest. Event though someone else's data may present challenges, such as duplicate or different keys as you pointed out. Which sectors companies get allocated to is also often useless and is better off being manually corrected.

At retail level, finding a consistent edge is not as much of a problem, as long as you stick to smaller, less liquid companies. At your end of town, however, how much of that development was ongoing? The edges that you found, how long would you normally have to exploit them before it was discovered by others?

Some other thoughts on challenges/considerations for this kind of software. My backtesting software, by the way, is not the only one of its kind I worked on. I also built software that analysed company financials for credit risk, which, under the bonnet, is almost the same thing. Funny you mention HiPortfolio, I used to work for them, although mainly worked on HiTrust.

- Data structure is super important. It may come from other sources, but if you don't make a copy and store it yourself, it will be slow. And then, it can be optimised for two purposes - speed of searching through all companies and speed of computation on one company. One will always be at expense of the other. The data will either be fast to search, or to compute. All relative, of course.

- Data will be BIG. And when running a multi year backtest, over thousands of companies, if you need to search the entire database to match the best possible trade, it will be slow. And so, highly denormalised data structure may be required, making it even bigger.

- Each company has many years of financials, each financial is made up of many figures, many computed. Some of those computed figures use current price as input. Which means they need to be recalculated every step.

- Database or flat file? It's 2014, yet this is one of the few cases where it is still a tough choice.

- survivorship bias is very important to guard against, results can be impacted very sustantially by it. And it's not just delisted companies. There may be companies that you have no interest investing in now, but would have been fine candidates 10 years ago. It needs to be possible to include these, but only to a certain point.

- how much data is enough? Would you trust backtest results if the strategy only resulted in 10 trades over 10 years in one market? What if all trades were in one year, with no matching opportunities in other 9? What if most trades were in one industry?

- stock splits, consolidations, rights issues are a pain in the ### to handle.

- there's an infinite number of data oddities that will screw up results. It is often better to just omit some companies from backtest, than to try and handle every combination.

- number of criteria is an important consideration. Designing a buy/sell criteria that looks at 20 different measures may result in a perfect result, but most likely means that you just fished for data that happened to work in that particular period. There's no good way to know whether that's the case or not. It's nice, as you pointed out before, that the criteria fits some plausible economic theory. Ideally it has been shown to work by other researchers in different markets at different times. The more researched it is, the less likely it is to produce above average results, however.

- combining two good backtest criteria into one does not always result in better strategy.

- timing issues. Record of results vs release of guidances, etc. This tends to be an issue for shorter term strategies. For strategies over many years, this tends to even out.

- selling criteria makes a lot less of a difference than buying criteria for long term strategies.

- when backtesting a scenario with limited funds, being fully invested or not, and when, makes a massive difference. Therefore, percentage of funds you allocate to each position, and how many positions you would normally expect to have becomes a big consideration.

- searching for what makes companies improvement revenue/earnings/etc. may be better than searching for ones that appreciate in price, although these usually overlap.

- doing backtests on some "common knowledge" makes you realise most of it is wrong.


Hopefully someone finds it useful or just interesting, I want to keep this thread alive.