This is a mobile optimized page that loads fast, if you want to load the real page, click this text.

Dump it Here

Jump to new


Mr Skate,

I found this post interesting. Why?

Because it somewhat addresses the issues of the 'Optimisation Paradox' and the often conflated 'Overfitting' issue.

So a definition of optimisation could be: parameter optimisation results in a system that is more likely to perform well in the future, but less likely to outperform the simulation.

Overfitting however is when the system becomes too (or overly) complex. The rules applied only effect a small number of trades that had outsized effects on the results.

The latter could apply to systems tested against short periods of historical data. Which circles back to an earlier discussion around the length of market historical data that should be used in testing a system. Obviously the correct answer is the more data the better, assuming that you want the system to trade for a long time.

Now you have used what has been described as a complex set of rules. Do you consider them (overly) complex?
The historical data is short. Too short?



What we are talking about is: long historical data - long gaps between high volatility and average/low volatility.
Short data run: overfitted for high volatility (or whatever market condition you are looking for).

Which (I'm guessing here a bit) would mean that if you used long data runs, the system would be (more) vulnerable to periods of high volatility. Whereas if you used a short data run, say from 2020-2023, where we had that high vol. event and higher average vol. for some time, assuming a return to high(er) vol. the system would (should) perform better?



Curious.

jog on
duc
 
ducati916 said:
I found this post interesting. Why?

Because it somewhat addresses the issues of the 'Optimisation Paradox' and the often conflated 'Overfitting' issue.
Click to expand...

@ducati916, thank you for raising an important topic around developing robust trading systems. You make excellent points about the risks of over-optimization and overfitting that can come from tweaking parameters solely to fit historical data.

I completely agree that optimization, rigorous out-of-sample testing, and favouring simpler logic tested across long time periods is the prudent approach. Regarding optimization, it's true that tweaking parameters to maximize past performance can lead to systems that fail to generalize well on new data.

On overfitting, systems can become overly complex and customised to limited historical periods. Simpler, robust logic tested across long timeframes is preferable.

Your note about being cautious about applying decades-old data to today's radically different trading landscape is also well-taken. More data is better, but markets evolve rapidly so solely relying on the past can be dangerous. There are certainly tradeoffs between optimization, complexity, data length, and recency that require careful balancing based on the specific strategy.

Basically, there are no easy answers here. Ultimately there are tradeoffs between optimization, complexity, and data usage that require experience and sound judgement. I aim to tread carefully in these areas, but appreciate you highlighting the subtleties involved.

Skate
 


The SAP Strategy was designed for those who may find system trading intimidating initially, with the aim of reducing the emotional stress of trading through selective entries and quick exits of non-performing trades. In my enthusiasm to bring a new system to the market, I regret that my haste led me to deploy it without thorough testing beforehand, for which I apologize.

The SAP Strategy uses a combination of technical indicators like moving averages, volatility, volume, momentum filters etc to identify high-probability buy setups that align with the overall market trend. To determine the trend, the strategy analyzes factors like trading volume, price momentum, and the percentage of stocks moving upward in a market index.

A key component is the "Percentage Up" filter which generates buy signals only when over 50% of the market index is in an uptrend, to avoid trades against the prevailing trend.

The strategy incorporates multiple exit rules like trailing stops, profit targets and time exits to control risk. Trades are exited swiftly if they fail to progress within a short period. The priority is on successful entries rather than speed while minimising losses.

Position sizing is calculated based on account balance and risk tolerance. The strategy also uses volatility-based stops that adapt to market conditions. The goal is to maximise returns while minimising risk.

It implements various tools like ATR, ADX, RSI, ROC, volume-weighted averages, etc. to time entries and exits. The objective is to weed out low-probability setups and identify high-confidence trades.

The SAP approach favours consistency over speed, with the focus being sustainable growth over the long term. While no strategy removes all risks, the SAP technique implements various tools like position sizing, stop losses, and volatility filters to control risk exposure. The goal is to provide a straightforward, methodical approach to develop a strategy with a low drawdown that has the potential for good returns

The strategy is designed for patience and discipline. It may miss some potential trades by waiting for optimal conditions but prefers consistency over speed.

Skate.
 

Playing the "Devils Advocate"
Based on the information I've already provided it does appear at face value the SAP Strategy incorporates a large number of indicators and filters. Here are some of my thoughts on how this could affect the strategy.

Pros
1. The "multiple indicators" help confirm trade signals, which can improve accuracy.
2. The use of "multiple filters" helps weed out lower probability setups and identify higher confidence trades.
3. The combination of "indicators and filters" aims to provide a robust multi-factor approach to finding good trades.

Cons
1. Too many indicators run the risk of curve-fitting and over-optimization of historical data rather than underlying market principles.
2. The large number of filters significantly reduces the number of trading opportunities. This could limit profit potential.
3. The combination of "indicators and filters" adds complexity without necessarily providing a meaningful edge over a strategy with fewer well-chosen indicators.

Overall, while my intent is valid, the high number of indicators does raise the concern of over-optimization by some members. I'm just saying (IMHO) the use of well-crafted indicators and filters improves the performance of the SAP Strategy enough to justify increased complexity. I also believe the strategy could be simplified without losing too much of its edge.

Starting with a few key indicators and expanding the mix cautiously was my initial goal in maximising robustness and simplicity. As time went on the main focus was to reduce drawdowns. In saying all this I'm content with how the SAP strategy turned out.

Skate.
 
ducati916 said:
Now you have used what has been described as a complex set of rules. Do you consider them (overly) complex?
Click to expand...

I'm curious about which filter or Indicators should be removed
Here is an expanded explanation of each indicator and filter used in the SAP trading strategy.

Hann Filter
The Hann filter removes noise and smoothes data. It helps identify trends and disregard insignificant price movements.

VWMA Filter
The Volume Weighted Moving Average gives more weight to periods with higher volume, as they are seen as more significant. It spots trends in price.

Elder Impulse Filter:
This clever filter Identifies acceleration and deceleration in trends. Helps spot trend direction changes.

Standalone Momentum Filter
The sole purpose of this filter is to identify individual positions that are building or losing momentum.

When it comes to Price, turnover, and volume filters have a read what @peter2 has to say:
https://www.aussiestockforums.com/t...her-asx-portfolio-wkly-dly.35039/post-1236364

Price Filter
This simple filter specifies the minimum and maximum price range for a position to be considered. The price filter ensures only positions within the set price range make it to the buy filter

Turnover Filter
This filter Specifies the minimum trading volume required for a stock to be considered, ensuring only actively traded stocks are bought.

Volume Filter
This filter specifies the minimum trading volume required, like a turnover filter to ensure only high-volume stocks are bought.

Rate of Change Momentum Filter
This handy little filter analyzes the pace of price change "over time" to find stocks gaining momentum which is important as other filters depend on the results of this filter.

Position Sizing
This criterion is to determine the amount invested in each stock so the allocates capital appropriately to handle the acceptable risk level.

Max Positions
This simply determines the maximum number of open positions. Prevents overexposure and adds a level to risk management.

Ranking System
Ranks buy signals from the lowest to the highest dollar value. My research indicates by doing so increases the profitability of a strategy.

Commission/Slippage
Account for trading cost variations.

Percentage Up Filter
Requires percentage of market trending up. Ensures broad upward momentum. This can cause delays in entering trends. Slow and steady wins the race analogy.

Skate.
 


A few SAP Strategy charts
The SAP Strategy has generated several buy and sell signals in the current calendar year (2023). In case a position fails to follow through, it is quickly sold. Similarly, when the take-profit stop is triggered, the position is sold to secure profits.









Skate.
 


Nothing is perfect
Even though I put a lot of research and testing into developing the SAP Strategy, it does not always work as intended when applied in the markets. No strategy can be perfect or right 100% of the time. Markets are dynamic and unpredictable, even what works with one position doesn't always transfer to the next. No mechanical system comes without flaws.

Individual stock performance is idiosyncratic as well
News and events can undermine the logic of a signal on a single position. False signals and whipsaws occur frequently even in strong trends. If I don't manage risk properly through smart stop loss placement, normal market fluctuations can force a position out early.

Trading is full of uncertainties
By not overtrading the SAP Strategy transaction costs don't keep chipping away at profits. While I aim for robustness in my strategy design, real-world performance will inevitably fall short of perfection due to inherent market uncertainties. This is the very reason why I need to keep optimising the strategy to manage risk more effectively and adapt to evolving market conditions. If you don't think the GFC, COVID, or the ongoing Russian/Ukraine War doesn't affect the market, I'm suggesting you think again.

At times the SAP Strategy disappoints
One loser after another.



Skate.
 
The only thing that I am certain of is

" The Good Get Gooder" and "The Bad Get Badder" "

Most of the Times!
 

Morning Mr Skate,

So Mr Curtis Faith (of Turtle Trader fame) ran a number of backtested strategies and found that ALL strategies performed better on ALL metrics when there were NO STOPS.

His conclusion: for TREND FOLLOWERS the losses come from the reversal of the trend. Not the volatility of the trend. These tend to be open profits (the drawdowns).

I suppose the question is (for anyone designing a mechanical system): what are you actually trading and in what timeframe?

So to provide some context: I trade a time period of 6mths to 1yr. I have no stops in the market. My exit is expiration of the Option. Those are my rules. The return = +/- 30%/annum. I will hold a mixture of CALLS/PUTS.

What is the common denominator? The ability to sit for long periods doing nothing and (endure) significant fluctuations (volatility) allowing the position to move.

The discussion for me with regard to your methodology, the pros/cons of so much complexity around the measurement of momentum/volatility and by inference, a stop or exit from the position.

To me your system is not a trend following system at all. It is a short term momentum system.

Which leads directly into a far more interesting discussion which is: Price Action.

Your current system is driven by price action. Your backtesting (in any period) is informed by price action. I ignore price action. I am seeking macro trends.

Price action is driven by many different market participants. Macro is driven by fundamentals. Fundamentals drive trends. Price action drives momentum.

So 2 charts:



Two examples of geopolitical macro trends driving price action and price action missing the big picture until it was slapped in the face (Bonds). The USD/Commodity correlation is far more interesting currently. I'll have more thoughts later.

Now of course, momentum based systems can trade the same trend on multiple entries/exits in a highly profitable manner.

What we are distinguishing is: time in the market v timing the market. Timing the market requires many more decisions, ie. added complexity. That added complexity is exemplified by your current list of entry/exit criteria.

My submission is: added complexity provides (many) more opportunities for failure. When there is more to go wrong, more will go wrong.

Anyway that was my third cup of coffee in this post...so enough.

I'll be interested on all or any thoughts.

jog on
duc
 
ducati916 said:
I'll be interested on all or any thoughts.
Click to expand...

@ducati916, although I do not trade options or the US markets, I appreciate your perspective. Your approach of holding options with no stops and relying on expiration as the exit is a valid strategy that appears to work well for you. It's crucial to find a trading system that aligns with your personality, risk tolerance, and timeframe.

Regarding the discussion on trend following versus momentum trading, there can be some overlap between the two. Momentum trading can be a subset of trend following, where traders look for short-term price movements in the direction of the overall trend. However, trend following typically involves holding positions for longer periods, and traders may use stops to manage risk.

Regarding the complexity of trading systems, it's true that adding more criteria and rules can increase the risk of failure. However, a well-designed system should be robust enough to handle a variety of market conditions and should have a sound theoretical basis. It's important to test any system thoroughly before trading it live and to be willing to adapt and modify the system as needed.

Every trader has their own set of strengths, weaknesses, preferences, and risk tolerances. What works for one trader may not work for another. Therefore, it is important for each trader to develop their own trading style and approach, taking into account their individual circumstances and characteristics. This involves experimenting with different strategies and techniques until finding the ones that work best for their goals and personality.

Maintaining discipline is crucial for successful trading. This includes sticking to a trading plan, avoiding emotional decisions, and managing risk effectively. Risk management is essential to preserve capital and avoid catastrophic losses. Traders should always consider the potential downside of trading and use appropriate risk-reducing strategies such as stop-loss orders or position sizing. Successful trading is all about managing risk.

Finally, the market is constantly changing, and traders must continuously learn and improve to remain successful. This involves staying up-to-date with market developments, analyzing past trades, and seeking out new information and insights. Successful traders are always looking for ways to improve their skills and strategies, and they are not afraid to adapt or modify their approach as needed.

Skate.
 
ducati916 said:
So Mr Curtis Faith (of Turtle Trader fame) ran a number of backtested strategies and found that ALL strategies performed better on ALL metrics when there were NO STOPS.
Click to expand...

Curtis Faith's discovery that backtested techniques perform better without stops is an intriguing one. However, it's crucial to remember that backtesting findings aren't necessarily predictive of real-world performance.

The effectiveness of any trading strategy is determined by a variety of factors, including the type of strategy you are trading, the timeframe used for backtesting, and the risk management measures used. It is feasible that a strategy without stops will perform better in some market conditions or periods but not in others.

ducati916 said:
His conclusion: for TREND FOLLOWERS the losses come from the reversal of the trend. Not the volatility of the trend. These tend to be open profits (the drawdowns).
Click to expand...

The remark that losses for trend followers are caused by the reversal of the trend rather than the volatility of the trend is correct to some degree. The fact that losses for trend followers are caused by trend reversal rather than trend volatility is significant. While volatility might impair the success of a trend-following strategy, it is true that trend reversals cause the most losses simply because traders are not quick enough to respond.

This is due to the fact that trend-following techniques often include holding positions for extended periods of time in order to capture the full range of the trend but doing so can result in large drawdowns, while the trader is still holding the position.

Overall, when developing a trend-following strategy, it's important to consider the risk of trend reversals and to have clear rules in place for exiting trades if the trend reverses. This can help to limit losses and protect the trader's capital over the long term.

Skate.
 
ducati916 said:
So to provide some context: I trade a time period of 6mths to 1yr. I have no stops in the market. My exit is expiration of the Option. Those are my rules. The return = +/- 30%/annum. I will hold a mixture of CALLS/PUTS.
Click to expand...
@ducati916 to be totally clear regarding your risk management, if you are buying calls/puts then you can only lose the amount of money that you paid for them. This means that you, in fact, do have hard stops in the market and therefore you are able to manage this risk.
 
DaveTrade said:
1. @ducati916 to be totally clear regarding your risk management, if you are buying calls/puts then you can only lose the amount of money that you paid for them.
Click to expand...

DaveTrade said:
2. This means that you, in fact, do have hard stops in the market and therefore you are able to manage this risk.
Click to expand...

Mr DaveTrade,

1. Correct.

2. In practical terms, sure I agree. However functionally, it is different. The position is only ever exited at expiry (win or lose) not as a function of volatility. Hard stops for stock/futures positions are subjected to volatility exiting the position. Of course the position can be re-entered.

Hard stops on stock/futures/cfd positions, when subjected to a 2010 flash-crash event (see video) can cost you many multiples of assumed risk.

Now I accept that the AUS market is different to the US market. However in the US stops are regularly (everyday) run by the market makers. Technical analysis is so predictable. Periodically they will run major levels with significant pullbacks. This makes placing a hard stop in the market difficult. Many will run mental stops. Fine if you are disciplined and can watch the market all day, everyday. A problem if you can't. Of course many position size based on the hard stop, another issue.

jog on
duc
 
DaveTrade said:
@ducati916 to be totally clear regarding your risk management
Click to expand...
ducati916 said:
position is only ever exited at expiry (win or lose) not as a function of volatility. Hard stops for stock/futures positions are subjected to volatility exiting the position.
Click to expand...

In the conversation between @DaveTrade and @ducati916, the Duc expressed concerns regarding risk management, specifically the limitations of hard stops in managing risk when trading options compared to traditional positions, especially in high volatility or flash crash situations. However, it's worth noting that when trading the ASX, there may be additional factors to consider that could impact risk management.

In this context, a multi-stage exit strategy, which involves a stale stop and an optional take profit stop, offers an alternative approach to managing risk and exiting positions. The stale stop is designed to prevent the trade from becoming stale and to exit the position if the market moves against the expected direction. The optional take profit stop involves using the ATR multiplied by a certain factor to set a target price for taking profits.

Overall, traders should consider the risks and limitations of their chosen trading strategy, including the effectiveness of hard stops and alternative approaches such as a multi-stage exit strategy. It is important to carefully backtest and evaluate the performance of any strategy before implementing it in live trading, and adjust the parameters as needed to fit individual trading styles and risk tolerances.

Skate.
 
This is the problem with trend trading in a nutshell
Trend-following strategies often hold positions for "extended periods of time" to capture the full range of a trend. However, holding a position for an extended period of time can result in large drawdowns. But I believe there is an answer to counteract this undesirable feature of trend trading.

Just to be clear
The purpose of trend trading is to catch a significant share of a trend. Trends, on the other hand, are not necessarily smooth and continuous, and they might undergo pullbacks, corrections, or even reversals. If the position is not managed appropriately, these movements can cause open profits to diminish or possibly turn into losses.

To manage this risk, traders can use a variety of techniques
One approach is to use a trailing stop, a stop that moves with the price as it moves in the trader's favour. This allows the trader to capture as much profit as possible while protecting against significant losses if the trend reverses.

Multi-stage exits
A stale exit on the other hand, which is a preset exit point based on price movement and time, is another approach. The trader can lock in profits and avoid handing back too much of their open profits if the trend reverses by using a stale exit.

Another method is to lock in profits
A take-profit exit is a predefined level at which a trader decides to exit the position and take the profit on offer. This enables the trader to exit at a predetermined level without losing too much of their open profits if the market reverses.

In conclusion
Managing open profits is a crucial component of successful trend trading. Traders can use a combination of trailing stops, stale exits, and take profit exits to manage their risk and capture as much profit as possible while protecting against significant losses. By doing so, traders can achieve their goal of capturing a significant share of a trend's movement without the need to hold positions for "extended periods of time".

Skate.
 
Trading can be a frustrating experience for many
For one, trading is inherently volatile, meaning share prices can fluctuate rapidly and unexpectedly, making it difficult to make informed decisions about when to buy and sell shares. Additionally, emotional attachment to positions can make it challenging to make rational decisions about when to sell.

Information overload
The vast amount of information available to traders can be overwhelming, and it can be difficult to determine which sources are reliable and relevant. Share trading is a competitive field, which can be frustrating for traders who feel like they're always one step behind. Trading always involves some level of risk, and traders are not always successful. When traders experience losses, it can be discouraging and frustrating.

In summary
Trading requires knowledge, discipline, and patience. It's important to have realistic expectations and to be prepared to face challenges and setbacks along the way.

Skate.
 
One small advantage of CFDs-on-futs over real futures is a guaranteed SL. It's expensive, and you have to place it a long way off the current price, but it's available.
 
@Skate








So the million dollar question: how to differentiate a trend reversal from simply volatility of the trend? How much vol. is too much vol.? When does a trend following strategy simply morph into buy and hold?

An Australian stock:



What we boil down to (again) is time in the market vs timing the market.

Timing the market calls for far more precision in entering/exiting the market. Far more complexity. You are, as you have described, taking little bites out of the trend on multiple entries/exits.

Those multiple entries/exits are (by definition) far more sensitive to small fluctuations in vol. This can have the effect of churning until a smooth trend develops. The cost of doing business.

Every now and then those small fluctuations in vol. become really major fluctuations in vol. When that happens, liquidity vanishes. Always. Your stops become (not quite) worthless and are executed (if at all) a significant distance from where planned. Possibly Mr Gringotts are the exception on CFDs.

My point: to win in this game (trend following) you need: (a) a high win rate or (b) winners to significantly outperform losers or (c) both.

The lower (a) the greater (b) must be or vice versa.

So my question would be by how much does the complexity or sensitivity of exits impact (b)?

Your exit strategy seems highly risk averse (not a bad thing in my book) and unlikely to catch a really huge trend. Is that by design?

jog on
duc
 
ducati916 said:
Your exit strategy seems highly risk averse (not a bad thing in my book) and unlikely to catch a really huge trend. Is that by design?
Click to expand...

@ducati916 you raise some very good points about the challenges of timing the market and differentiating between volatility and trend reversals.

There are a few things I would note
(1) No exit strategy will be perfect. There will always be some trades that exit too early or too late. The key is finding a strategy that works well overall across many trades. Some volatility and imperfection have to be accepted.

(2) My exit strategy is intentionally designed to be risk-averse, as you noted. The goal is to capture solid gains while limiting losses. It sacrifices some potential upside in order to better protect the downside. This is a common tradeoff in trading system design.

(3) The exact parameters like moving average periods, indicator thresholds, etc. can be optimised to balance trading frequency, win rate, average gain/loss, max drawdown, and other metrics. So there is room to fine-tune the strategy for different goals.

(4) No exit will fully avoid slippage in fast markets. But using a basket of indicators rather than just a hard stop can help smooth out exits across multiple bars and help avoid whipsaws.

(5) Trend following often underperforms in choppy sideways markets and outperforms in strong trending markets. So market conditions play a big role in results.

(6) Even with best efforts, there will be times when volatility picks up right as you enter, leading to an early exit. Mistiming is inevitable sometimes.

Skate.
 
ducati916 said:
So my question would be by how much does the complexity or sensitivity of exits impact
Click to expand...

Let's talk about Churning (entering and exiting trades quickly without large gains)
Some amount of churn is inevitable and must be accepted as a cost of actively trading strategies with "sensitive exits". The key is evaluating if the strategy's churn is excessive compared to the profitable trends it captures. If there is too much churn, revisiting the rules to avoid over-trading is needed.

The goal is to find the right balance
There is a constant balance between having responsive exits to limit losses versus avoiding exiting too early before major trends. No exit strategy will be perfect, so the goal should be continual improvement while evaluating performance over many trades and diverse market conditions. With refinement over time, a well-balanced exit strategy can be developed that protects capital during corrections without missing out on capturing large trend-based gains.

Responsive exits are the key to trading profitably
In general, managing churn is a key consideration when developing and optimising an exit strategy. The key is evaluating results over many trades and across various market conditions. There are always ways to try and improve an exit strategy but it will never be perfect.

Skate.
 
You must log in or register to reply here.

Similar threads

R
My strategy/plan going forward
Replies
19
Views
1K
divs4ever
Very nice Rain here today
4 5 6
Replies
105
Views
5K
farmerge
  • Poll
Financial Independence (Home Ownership, Super, Long Retirement) not possible for wage earners: What do we need and how do we get it?
11 12 13
Replies
258
Views
21K
Smurf1976
November 2024 DDD
2 3
Replies
53
Views
6K
ducati916
Ozempic: Magic Pill or Devils Bargain?
Replies
10
Views
2K
sptrawler
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more...