C'mon @ducati916
| Small-caps are stepping into the spotlight after sitting on the sidelines for a long stretch. These stocks have been waiting for their moment, and it looks like that moment is finally here. The Russell 2000 ETF just closed at new all-time highs for the first time in 969 days — ending its longest drought ever. In fact, the Russell 2000 ETF finally closed at new all-time highs for the first time in 969 days — ending its longest drought ever. |
 |
Now it’s their turn to go on offense and prove they can sustain this strength. If small-caps can keep leading from here, it would be a huge confirmation of risk appetite and a fresh tailwind for the broader market into the year-end. A lot of these names are already participating — Industrials, Tech, Gold miners. These are offensive groups pushing higher. And really, it’s tough to picture this bull market wrapping up before small-caps even register a clean breakout. That’s been a key reason we’ve stayed constructive and eager to keep putting money to work in equities. Our Chief Market Strategist Steve Strazza walked through this yesterday in a live session, highlighting actionable setups. One standout: the 2-to-100 Hunt. |
 |
Here's another one from Yahoo! Finance:
Then there was a research note from Torsten Slok, the chief economist at private-equity and alternative asset management firm Apollo Global Management:
It went on to assert that the top 10 companies in the S&P 500 today are more overvalued than they were back then.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| We start with the best sectors, then drill into the subgroups. We pick one, and then take a look at the top stocks in it. This week’s standout is Energy, which just jumped to the second spot in our sector rankings. |
 |
| Energy was the best-performing sector this week, with $XLE reaching its highest level since early April. Beneath the surface, many individual stocks are breaking out of short-term consolidation patterns. This suggests we could see some follow-through leadership from energy in the coming weeks. Here’s a look at our overall industry rankings, which show Energy Equipmentstocks reaching the sixth spot. |
 |
| This group supplies the machinery, tools, and services that energy producers rely on to explore for and extract oil and gas. The relative strength from equipment names has been off the charts of late. This is a risk-on signal for the broader energy trade. Below are the Top 10 names in the Energy Equipment subsector, ranked by relative strength. |
 |
| This week’s spotlight stock is Solaris Energy Infrastructure $SEI: |
 |
| The stock carries a massive 30% short interest with a 5x days-to-cover ratio as price presses against the upper bounds of a one-year base. |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
, But Common Mate, it's going to take me a week to read all these posts!
Welcome Back Duc
| Chart of the Day for September 29, 2025 |
|
Today’s Featured Stock |
| Valued at $4.61 billion, Mercury General (MCY) is engaged primarily in writing all risk classifications of automobile insurance. The company offers automobile policyholders the following types of coverage: bodily injury liability, underinsured and uninsured motorist, property damage liability, comprehensive, and collision. |
What I’m Watching |
| I found today’s Chart of the Day by using Barchart’s powerful screening functions to sort for stocks with the highest technical buy signals; superior current momentum in both strength and direction; and a Trend Seeker “buy” signal. I then used Barchart’s Flipcharts feature to review the charts for consistent price appreciation. MCY checks those boxes. Since the Trend Seeker signaled a "Buy" on Aug. 4 the stock gained 16.07%. |
 |
Barchart Technical Indicators for Mercury General |
| Editor’s Note: The technical indicators below are updated live during the session every 20 minutes and can therefore change each day as the market fluctuates. The indicator numbers shown below therefore may not match what you see live on the Barchart.com website when you read this report. These technical indicators form the Barchart Opinion on a particular stock. |
Mercury hit an all-time high of $83.34 in morning trading on Sept. 29.
|
Don’t Forget the Fundamentals |
|
Analyst and Investor Sentiment on Mercury General |
| I don’t buy stocks because everyone else is buying, but I do realize that if major firms and investors are dumping stock, it’s hard to make money swimming against the tide. It looks like Wall Street analysts are high on MCY and so are individual investors.
|
The Bottom Line on Mercury General |
| MCY appears to have backing not only from Wall Street, but also from other financial advisory sites and individual investors. |
|
|
|
|
|
|
A long long time ago, while studying for my engineering degree with a specialisation in AI, we worked with neural networks.Now this is for the traders out there that love complicated:
View attachment 209789
Hello everyone,
It has been a while since I posted an indicator, so thought I would share this project I did for fun.
This indicator is an attempt to develop a pseudo Random Forest classification decision matrix model for Pinescript.
This is not a full, robust Random Forest model by any stretch of the imagination, but it is a good way to showcase how decision matrices can be applied to trading and within Pinescript.
As to not market this as something it is not, I am simply calling it the "Simple Decision Matrix Classification Algorithm". However, I have stolen most of the aspects of this machine learning algo from concepts of Random Forest modelling.
How it works:
With models like Support Vector Machines (SVM), Random Forest (RF) and Gradient Boosted Machine Learning (GBM), which are commonly used in Machine Learning Classification Tasks (MLCTs), this model operates similarity to the basic concepts shared amongst those modelling types. While it is not very similar to SVM, it is very similar to RF and GBM, in that it uses a "voting" system.
What do I mean by voting system?
How most classification MLAs work is by feeding an input dataset to an algorithm. The algorithm sorts this data, categorizes it, then introduces something called a confusion matrix (essentially sorting the data in no apparently order as to prevent over-fitting and introduce "confusion" to the algorithm to ensure that it is not just following a trend).
From there, the data is called upon based on current data inputs (so say we are using RSI and Z-Score, the current RSI and Z-Score is compared against other RSI's and Z-Scores that the model has saved). The model will process this information and each "tree" or "node" will vote. Then a cumulative overall vote is casted.
How does this MLA work?
This model accepts 2 independent variables. In order to keep things simple, this model was kept as a three node model. This means that there are 3 separate votes that go in to get the result. A vote is casted for each of the two independent variables and then a cumulative vote is casted for the overall verdict (the result of the model's prediction).
The model actually displays this system diagrammatically and it will likely be easier to understand if we look at the diagram to ground the example:
In the diagram, at the very top we have the classification variable that we are trying to predict. In this case, we are trying to predict whether there will be a breakout/breakdown outside of the normal ATR range (this is either yes or no question, hence a classification task).
So the question forms the basis of the input. The model will track at which points the ATR range is exceeded to the upside or downside, as well as the other variables that we wish to use to predict these exceedences. The ATR range forms the basis of all the data flowing into the model.
Then, at the second level, you will see we are using Z-Score and RSI to predict these breaks. The circle will change colour according to "feature importance". Feature importance basically just means that the indicator has a strong impact on the outcome. The stronger the importance, the more green it will be, the weaker, the more red it will be.
We can see both RSI and Z-Score are green and thus we can say they are strong options for predicting a breakout/breakdown.
So then we move down to the actual voting mechanisms. You will see the 2 pink boxes. These are the first lines of voting. What is happening here is the model is identifying the instances that are most similar and whether the classification task we have assigned (remember out ATR exceedance classifier) was either true or false based on RSI and Z-Score.
These are our 2 nodes. They both cast an individual vote. You will see in this case, both cast a vote of 1. The options are either 1 or 0. A vote of 1 means "Yes" or "Breakout likely".
However, this is not the only voting the model does. The model does one final vote based on the 2 votes. This is shown in the purple box. We can see the final vote and result at the end with the orange circle. It is 1 which means a range exceedance is anticipated and the most likely outcome.
The Data Table Component
The model has many moving parts. I have tried to represent the pivotal functions diagrammatically, but some other important aspects and background information must be obtained from the companion data table.
If we bring back our diagram from above:
We can see the data table to the left.
The data table contains 2 sections, one for each independent variable. In this case, our independent variables are RSI and Z-Score.
The data table will provide you with specifics about the independent variables, as well as about the model accuracy and outcome.
If we take a look at the first row, it simply indicates which independent variable it is looking at. If we go down to the next row where it reads "Weighted Impact", we can see a corresponding percent. The "weighted impact" is the amount of representation each independent variable has within the voting scheme. So in this case, we can see its pretty equal, 45% and 55%, This tells us that there is a slight higher representation of z-score than RSI but nothing to worry about.
If there was a major over-respresentation of greater than 30 or 40%, then the model would risk being skewed and voting too heavily in favour of 1 variable over the other.
If we move down from there we will see the next row reads "independent accuracy". The voting of each independent variable's accuracy is considered separately. This is one way we can determine feature importance, by seeing how well one feature augments the accuracy. In this case, we can see that RSI has the greatest importance, with an accuracy of around 87% at predicting breakouts. That makes sense as RSI is a momentum based oscillator.
Then if we move down one more, we will see what each independent feature (node) has voted for. In this case, both RSI and Z-Score voted for 1 (Breakout in our case).
You can weigh these in collaboration, but its always important to look at the final verdict of the model, which if we move down, we can see the "Model prediction" which is "Bullish".
If you are using the ATR breakout, the model cannot distinguish between "Bullish" or "Bearish", must that a "Breakout" is likely, either bearish or bullish. However, for the other classification tasks this model can do, the results are either Bullish or Bearish.
Using the Function:
Okay so now that all that technical stuff is out of the way, let's get into using the function. First of all this function innately provides you with 3 possible classification tasks. These include:
1. Predicting Red or Green Candle
2. Predicting Bullish / Bearish ATR
3. Predicting a Breakout from the ATR range
The possible independent variables include:
1. Stochastics,
2. MFI,
3. RSI,
4. Z-Score,
5. EMAs,
6. SMAs,
7. Volume
The model can only accept 2 independent variables, to operate within the computation time limits for pine execution.
Let's quickly go over what the numbers in the diagram mean:
The numbers being pointed at with the yellow arrows represent the cases the model is sorting and voting on. These are the most identical cases and are serving as the voting foundation for the model.
The numbers being pointed at with the pink candle is the voting results.
Extrapolating the functions (For Pine Developers:
So this is more of a feature application, so feel free to customize it to your liking and add additional inputs. But here are some key important considerations if you wish to apply this within your own code:
1. This is a BINARY classification task. The prediction must either be 0 or 1.
2. The function consists of 3 separate functions, the 2 first functions serve to build the confusion matrix and then the final "random_forest" function serves to perform the computations. You will need all 3 functions for implementation.
3. The model can only accept 2 independent variables.
I believe that is the function. Hopefully this wasn't too confusing, it is very statsy, but its a fun function for me! I use Random Forest excessively in R and always like to try to convert R things to Pinescript.
Hope you enjoy!
LOL.
Now for the money shot!
Pine Script®
// This source code is subject to the terms of the Mozilla Public License 2.0 at https://mozilla.org/MPL/2.0/
// /$$$$$$ /$$ /$$
// /$$__ $$ | $$ | $$
//| $$ \__//$$$$$$ /$$$$$$ /$$ /$$ /$$$$$$ /$$$$$$ /$$$$$$$ /$$$$$$ /$$$$$$ /$$ /$$ /$$$$$$ /$$$$$$$
//| $$$$$$|_ $$_/ /$$__ $$| $$ /$$//$$__ $$ /$$__ $$ /$$_____/|_ $$_/ /$$__ $$| $$ /$$//$$__ $$ /$$_____/
// \____ $$ | $$ | $$$$$$$$ \ $$/$$/| $$$$$$$$| $$ \__/| $$$$$$ | $$ | $$$$$$$$ \ $$/$$/| $$$$$$$$| $$$$$$
// /$$ \ $$ | $$ /$$| $$_____/ \ $$$/ | $$_____/| $$ \____ $$ | $$ /$$| $$_____/ \ $$$/ | $$_____/ \____ $$
//| $$$$$$/ | $$$$/| $$$$$$$ \ $/ | $$$$$$$| $$ /$$$$$$$/ | $$$$/| $$$$$$$ \ $/ | $$$$$$$ /$$$$$$$/
// \______/ \___/ \_______/ \_/ \_______/|__/ |_______/ \___/ \_______/ \_/ \_______/|_______/
//@version=6
indicator("Simple Decision Matrix Classification Algorithm [SS]", shorttitle = "DMCA [SS]", overlay=false)
g1 = "Model Parameters",
g2 = "EMA"
g3 = "RSI"
g4 = "MFI"
g5 = "Z-Score"
g6 = "ATR"
training = input.int(850, "Training Length", group = g1)
select_x1 = input.string("Stochastic", "Select Independent Variable X1", ["Stochastic", "RSI", "Z-Score", "MFI", "Volume", "EMA", "SMA"], group = g1)
select_x2 = input.string("Volume", "Select Independent Variable X1", ["Stochastic", "RSI", "Z-Score", "MFI", "Volume", "EMA", "SMA"], group = g1)
select_y = input.string("Red/Green Candle", "Select desired target Classification Prediction", ["Red/Green Candle", "Bullish / Bearish ATR", "ATR Range Breakout"], group = g1)
showtable = input.bool(true, "Show Table", group = g1)
// EMA
emalen = input.int(14, "EMA/SMA Length", group = g2), emasrc = input.source(close, "EMA/ SMA Source", group = g2)
// RSI
rsilen = input.int(14, "RSI Length", group = g3), rsisrc = input.source(close, "RSI Source", group = g3)
// MFI
mfilen = input.int(14, "MFI Length", group = g4), mfisrc = input.source(hlc3, "MFI Source", group = g4)
// Z-Score
zlen = input.int(14, "Z-Score Length", group = g5), zsrc = input.source(close, "Z-Score Source", group = g5)
// atr
atrlen = input.int(14, "ATR Length", group = g6)
atr = ta.atr(atrlen)
sent = close > open ? 1 : 0
sentiment = high - open > open - low ? 1 : 0
rsi = ta.rsi(rsisrc, rsilen)
sto = ta.stoch(close, high, low, 14)
mfi = ta.mfi(mfisrc, mfilen)
z = (zsrc - ta.sma(zsrc,zlen)) / ta.stdev(zsrc, zlen)
ema = ta.ema(emasrc, emalen)
sma = ta.sma(emasrc, emalen)
atr_breakout = high-open > atr or open-low < atr ? 1 : 0
float x1_input = switch select_x1
"Stochastic" => sto
"RSI" => rsi
"Z-Score" => z
"MFI" => mfi
"Volume" => volume
"EMA" => close - ema
"SMA" => close - sma
float x2_input = switch select_x2
"Stochastic" => sto
"RSI" => rsi
"Z-Score" => z
"MFI" => mfi
"Volume" => volume
"EMA" => close - ema
"SMA" => close - sma
float y_input = switch select_y
"Red/Green Candle" => sent
"Bullish / Bearish ATR" => sentiment
"ATR Range Breakout" => atr_breakout
float x1_threshold = 0
float x2_threshold = 0
if select_x1 == "Volume"
x1_threshold := ta.stdev(volume, 14)
else if select_x1 == "Z-Score"
x1_threshold := 0.05
else
x1_threshold := 0.5
if select_x2 == "Volume"
x2_threshold := ta.stdev(volume, 14)
else if select_x2 == "Z-Score"
x2_threshold := 0.05
else
x2_threshold := 0.5
// Function to shuffle a list of indices
shuffle_indices(n) =>
indices = array.new_int(n)
for i = 0 to n - 1
array.set(indices, i, i)
for i = 0 to n - 1
rand_index = math.floor(math.random(0, n))
temp = array.get(indices, i)
array.set(indices, i, array.get(indices, rand_index))
array.set(indices, rand_index, temp)
indices
// Function to shuffle the rows of a matrix
shuffle_matrix(matrix, rows) =>
shuffled_matrix = matrix.new<float>(rows, matrix.columns(matrix), 0)
indices = shuffle_indices(rows)
for i = 0 to rows - 1
for j = 0 to matrix.columns(matrix) - 1
matrix.set(shuffled_matrix, i, j, matrix.get(matrix, array.get(indices, i), j))
shuffled_matrix
radnom_forest_classification(x1, x2, y, train) =>
x1m = matrix.new<float>(train , 2, 0)
x2m = matrix.new<float>(train, 2, 0)
lx1 = x1[1]
lx2 = x2[1]
ly = y
for i = 0 to train - 1
matrix.set(x1m, i, 0, lx1)
matrix.set(x1m, i, 1, ly)
matrix.set(x2m, i, 0, lx2)
matrix.set(x2m, i, 1, ly)
x1m_random = shuffle_matrix(x1m, train)
x2m_random = shuffle_matrix(x2m, train)
int y1_pass = 0
int y1_fail = 0
for i = 0 to train - 2
if matrix.get(x1m_random, i, 0) > 0
if matrix.get(x1m_random, i, 0) >= x1 - x1_threshold and matrix.get(x1m_random, i, 0) <= x1 + x1_threshold
if matrix.get(x1m_random, i, 1) == 1
y1_pass += 1
else if matrix.get(x1m_random, i, 1) != 1
y1_fail += 1
int y2_pass = 0
int y2_fail = 0
int x1_vote = y1_pass > y1_fail ? 1 : 0
for i = 0 to train - 2
if matrix.get(x2m_random, i, 0) >= x2 - x2_threshold and matrix.get(x2m_random, i, 0) <= x2 + x2_threshold
if matrix.get(x2m_random, i, 1) == 1
y2_pass += 1
else if matrix.get(x2m_random, i, 1) != 1
y2_fail += 1
passes = y1_pass + y2_pass
fails = y1_fail + y2_fail
vote = passes > fails ? 1 : 0
x2_vote = y2_pass > y2_fail ? 1 : 0
// X1 independent accuracy
int x1_ia_pass = 0
int x1_ia_fail = 0
int x2_ia_pass = 0
int x2_ia_fail = 0
x2_vote_pass = x2_vote == y
x1_vote_pass = x1_vote == y
for i = 0 to train
if x1_vote_pass
x1_ia_pass += 1
else
x1_ia_fail += 1
if x2_vote_pass
x2_ia_pass += 1
else
x2_ia_fail += 1
x1_accuracy = x1_ia_pass / (x1_ia_pass + x1_ia_fail) * 100
x2_accuracy = x2_ia_pass / (x2_ia_pass + x2_ia_fail) * 100
// Accuracy
int success = 0
int failure = 0
for i = 0 to train - 1
if vote == y
success += 1
else if vote != y
failure += 1
accurcay = success / (success + failure) * 100
[vote, accurcay, y1_pass, y1_fail, y2_pass, y2_fail, x1_vote, x2_vote, x1_accuracy, x2_accuracy]
[result, acc, y1p, y1f, y2p, y2f, x1v, x2v, x1a, x2a] = radnom_forest_classification(x1_input, x2_input, y_input, training)
// Demographics
total_sorted_cases_x1 = y1p + y1f
total_sorted_cases_x2 = y2p + y2f
x1_voted_result = y1p > y1f ? 1 : 0
x2_voted_result = y2p > y2f ? 1 : 0
total_cases = total_sorted_cases_x1 + total_sorted_cases_x2
x1_total_weight = (total_sorted_cases_x1 / total_cases) * 100
x2_total_weight = (total_sorted_cases_x2 / total_cases) * 100
string_Result = result == 1 ? "Bullish" : "Bearish"
if barstate.islast
for lin in line.all
line.delete(lin)
for txt in label.all
label.delete(txt)
// Feature Importance
feature_importance_color(current_variable, other_variable) =>
color feature_color = na
if current_variable > other_variable
if current_variable <= 45
feature_color := color.rgb(5, 46, 7)
else if current_variable > 45 and current_variable <= 55
feature_color := color.rgb(10, 99, 14)
else if current_variable > 55
feature_color := color.rgb(5, 235, 15)
else if other_variable > current_variable
if current_variable > 50 and current_variable <= 55
feature_color := color.rgb(22, 110, 26)
else if current_variable > 55
feature_color := color.rgb(5, 235, 15)
else if current_variable <= 50 and current_variable > 45
feature_color := color.rgb(89, 6, 31)
else if current_variable <= 45
feature_color := color.rgb(247, 0, 75)
feature_color
if barstate.islast
label.new(bar_index - 200, y = 20, text = str.tostring(select_x1), color = feature_importance_color(x1a, x2a), style = label.style_circle, size = size.large)
label.new(bar_index - 250, y = 10, text = str.tostring(y1p), color = color.lime, style = label.style_circle, size = size.large)
label.new(bar_index - 150, y = 10, text = str.tostring(y1f), color = color.red, style = label.style_circle, size = size.large)
label.new(bar_index - 200, y = 0, text = str.tostring(x1v), color = color.purple, style = label.style_circle, size = size.large)
line.new(bar_index - 200, y1 = 20, x2 = bar_index - 250, y2 = 10, color = color.gray, width = 3)
line.new(bar_index - 250, y1 = 10, x2 = bar_index - 200, y2 = 0, color = color.gray, width = 3)
line.new(bar_index - 200, y1 = 20, x2 = bar_index - 150, y2 = 10, color = color.gray, width = 3)
line.new(bar_index - 150, y1 = 10, x2 = bar_index - 200, y2 = 0, color = color.gray, width = 3)
line.new(bar_index, y1 = 20, x2 = bar_index - 50, y2 = 10, color = color.gray, width = 3)
line.new(bar_index - 50, y1 = 10, x2 = bar_index, y2 = 0, color = color.gray, width = 3)
line.new(bar_index, y1 = 20, x2 = bar_index + 50, y2 = 10, color = color.gray, width = 3)
line.new(bar_index + 50, y1 = 10, x2 = bar_index, y2 = 0, color = color.gray, width = 3)
label.new(bar_index, y = 20, text = str.tostring(select_x2), color = feature_importance_color(x2a, x1a), style = label.style_circle, size = size.large)
label.new(bar_index - 50, y = 10, text = str.tostring(y2p), color = color.lime, style = label.style_circle, size = size.large)
label.new(bar_index + 50, y = 10, text = str.tostring(y2f), color = color.red, style = label.style_circle, size = size.large)
label.new(bar_index, y = 0, text = str.tostring(x2v), color = color.purple, style = label.style_circle, size = size.large)
label.new(bar_index - 100, y = - 10, text = str.tostring(result), color = color.orange, style = label.style_circle, size = size.large)
line.new(bar_index - 200, y1 = 0, x2 = bar_index - 100, y2 = -10, color = color.gray, width = 3)
line.new(bar_index, y1 = 0, x2 = bar_index - 100, y2 = -10, color = color.gray, width = 3)
label.new(bar_index - 100, y = 30, text = str.tostring(select_y), color = color.navy, style = label.style_circle, size = size.large)
line.new(bar_index - 100, y1 = 30, x2 = bar_index - 200, y2 = 20, color = color.gray, width = 3)
line.new(bar_index - 100, y1 = 30, x2 = bar_index, y2 = 20, color = color.gray, width = 3)
data = table.new(position.middle_left, 10, 10, bgcolor = color.rgb(0, 0, 0), frame_color = color.white, frame_width = 3)
if showtable
table.merge_cells(data, 1, 0, 2, 0)
table.merge_cells(data, 3, 0, 4, 0)
table.cell(data, 1, 0, text = str.tostring(select_x1), bgcolor = color.new(color.gray, 75), text_color = color.white, text_halign = text.align_center)
table.cell(data, 3, 0, text = str.tostring(select_x2), bgcolor = color.new(color.gray, 75), text_color = color.white, text_halign = text.align_center)
table.cell(data, 1, 1, text = "X1 Independent: ", text_color = color.white)
table.cell(data, 2, 1, text = str.tostring(select_x1), text_color = color.white)
table.cell(data, 1, 2, text = "X1 Total Sorted Cases: ", text_color = color.white)
table.cell(data, 2, 2, text = str.tostring(total_sorted_cases_x1), text_color = color.white)
table.cell(data, 1, 2, text = "X1 Weighted Impact: ", text_color = color.white)
table.cell(data, 2, 2, text = str.tostring(math.round(x1_total_weight, 2)), text_color = color.white)
table.cell(data, 1, 3, text = "X1 Independent Accuracy: ", text_color = color.white)
table.cell(data, 2, 3, text = str.tostring(math.round(x1a, 2)) + "%", text_color = color.white)
table.cell(data, 1, 4, text = "X1 Voted Result: ", text_color = color.white)
table.cell(data, 2, 4, text = str.tostring(x1_voted_result), text_color = color.white)
table.cell(data, 3, 1, text = "X2 Independent: ", text_color = color.white)
table.cell(data, 4, 1, text = str.tostring(select_x2), text_color = color.white)
table.cell(data, 3, 2, text = "X2 Total Sorted Cases: ", text_color = color.white)
table.cell(data, 4, 2, text = str.tostring(total_sorted_cases_x2), text_color = color.white)
table.cell(data, 3, 2, text = "X2 Weighted Impact: ", text_color = color.white)
table.cell(data, 4, 2, text = str.tostring(math.round(x2_total_weight, 2)), text_color = color.white)
table.cell(data, 3, 3, text = "X2 Independent Accuracy: ", text_color = color.white)
table.cell(data, 4, 3, text = str.tostring(math.round(x2a, 2)) + "%", text_color = color.white)
table.cell(data, 3, 4, text = "X2 Voted Result: ", text_color = color.white)
table.cell(data, 4, 4, text = str.tostring(x2_voted_result), text_color = color.white)
table.merge_cells(data, 1, 5, 4, 5)
table.cell(data, 1, 5, text = "Model Accuracy: " + str.tostring(math.round(acc, 2)), text_color = color.white, text_halign = text.align_center)
table.merge_cells(data, 1, 6, 4, 6)
table.cell(data, 1, 6, text = "Model Prediction: " + str.tostring(string_Result), text_color = color.white, text_halign = text.align_center)
WTF.
I'm hoping @qldfrog can make something of this.
jog on
duc
When the U.S. government heads toward a partial shutdown, a natural instinct is to anticipate dire consequences for financial markets and the economy.
|
| Newly published plans from the Department of Labor show that a shutdown this week will delay critical economic releases — Friday's jobs report and more — at a critical moment for the Federal Reserve. Why it matters: If lawmakers fail to strike a deal, a shutdown could drag on for weeks — postponing the publication of jobs and inflation indicators at a time when the Fed is considering how much to cut rates with sticky inflation and weak hiring. Flashback: The Fed was grappling with the potential of a debt ceiling breach during the 2013 partial government shutdown. That's not the case this time.
The other side: At the meeting, New York Fed president John Williams said the shutdown "complicated our job ... but we are far from flying blind," acknowledging private-sector data that helped fill the information gap.
|
Hedge funds and prop trading companies are going to be increasingly competing not just for entry-level talent — freshly-graduated computer scientists, mathematicians and general brain boxes — but also for mid-career people. Millennium’s fateful poaching of two Jane Street traders is therefore probably just the beginning. We’re going to see more of this happening — given that some hedge funds can pony up $100mn packages for portfolio managers — but we’re also likely going to see quant hedge fund people migrate to the prop trading world. After all, even the interns make stupid money. 
Prop trading firms are becoming increasingly important clients of Wall Street, and this is just going to become even more pronounced. On the current trajectory they soon might rival hedge funds and private equity for importance. Historically, prop trading firms operated fairly separately from the banks.
Prop trading firms are going to be raising more external capital, and hedge funds will become increasingly tempted to hive off their best trading strategies in internal funds for their own partners and employees. In fact, both of these things are already happening — at least in some form. Citadel Securities, Hudson River Trading and Jane Street have all tapped the debt markets to boost the firepower offered by their retained profits. Given how much money these firms have made in recent years they may never want or need to start more traditional fund-like investment vehicles, but Tower Research — one of the HFT pioneers — has talked to investors about doing so, and several industry insiders predict that “proper” HFT-powered long-short equity funds will inevitably emerge. At the same time, some of the higher-profile star-arb hedge funds are either already entirely employee money (Renaissance’s Medallion) or probably heading in that direction (DE Shaw’s Valence fund). It’s natural that more and more successful hedge funds start housing their low-capacity, high-Sharpe quasi prop-trading strategies in internal funds — even if it annoys investors no end and needs to be done very carefully.
Increasing competition in mid-frequency trading — generally said to be in the 1-5 day range — could cause crowding in some signals. The dangers of this crowding are ramped up by the growing use of leverage to maximise profits. There are good reasons to be sanguine about this. A more diverse set of participants is generally a healthy thing for a market. Prop trading firms overwhelmingly deploy their own money — the stickiest capital there is. Both prop firms and quant hedge funds are as a rule pretty obsessive about risk, and particularly assiduous about monitoring for signs of herd-like behaviour. However, as the head of one big trading firm told Alphaville: It is a concern how much crowding there is in stat arb signals right now. July was a sign that things could be very crowded . . . It’s unclear how much of it is due to this, but there’s a lot of money chasing similar strategies and signals in similar instruments, which can cause correlated drawdowns. Where on the Richter scale the next quant quake will measure is beyond Alphaville. But we can see that having nearly two-thirds of US equity trading volume — or ca 20x the total trading volume of the entire long-only investment fund industry — potentially getting locked into correlated drawdowns might not be ideal.
i meant ticker/exchange code
Hello and welcome to Aussie Stock Forums!
To gain full access you must register. Registration is free and takes only a few seconds to complete.
Already a member? Log in here.
