ask Dcifrths...well, anything...about finance/mkts/ports that is.

[silentbob]

Earlier this year, I moved a huge chunk of my 401(k) to an "inflation-protected securities" fund in a simplistic desire to escape equities, without knowing what the fund actually was. Yeah, not too smart.

So...having luckboxed into a nice return thus far, I'm wondering:

1) Why have inflation-protected securities done so well this year?

2) Which economic indicator(s) would portend a downturn in their value?

I tried reading the Wikipedia entry on TIPS and couldn't get past the first few sentences.

Thanks for all your great insights in this thread!

[crunchi]

Dcifrths,

Would you mind sharing some of the online resources that would be beneficial to noobs and Pros who want to further their understanding of investments, markets etc

[kimchi]

Edit Cliif's notes:- measuring risk adjusted performance easy calc?

Barron, excuse any intoxicated ramblings here, but I am always obsessed withthe risk I'm exposed to, and i think you have a pretty good grasp of quantifying risk. I have most of my money split 3 ways; buy and hold, a long term trend following strategy (averages 1 trade per year per market) and shorter term swing trading (averages 1 trade per month per market - although my most dominant position here is 'stand aside').

I have data for these (active strategies) going back a couple of years and I've back-teated them to several decades. They all 'beat the market' but at increased risk. I calculate risk daily to help me decide on position size using ATR, but I'd like to calculate risk in comparison to some random index (such as MCSI Worlsd or other diversified index).

What's the best way to do this? Sharpe ratio? Stddev? I remember reading a post of your a few weeks back where you talked about measuring risk-adjusted return in some way but I'm a lemon with the serach facility.

If you have monthly, hourly, or whatever results from your own methodology, what's the best way to compare your actice results with those of a passive index investor or the like?

I'd hate to think I could have just replicated my last 5 years worth of results by leveraging the FTSE All-Share with an IMR of 50% with substantially less risk.

I've tested and (small scale) traded various strategies but I received 16% for my uni biometrics (stats) exam and will have to stop my scotch rambling.

[DcifrThs]

[ QUOTE ]
Edit Cliif's notes:- measuring risk adjusted performance easy calc?

Barron, excuse any intoxicated ramblings here, but I am always obsessed withthe risk I'm exposed to, and i think you have a pretty good grasp of quantifying risk. I have most of my money split 3 ways; buy and hold, a long term trend following strategy (averages 1 trade per year per market) and shorter term swing trading (averages 1 trade per month per market - although my most dominant position here is 'stand aside').

I have data for these (active strategies) going back a couple of years and I've back-teated them to several decades. They all 'beat the market' but at increased risk. I calculate risk daily to help me decide on position size using ATR, but I'd like to calculate risk in comparison to some random index (such as MCSI Worlsd or other diversified index).

What's the best way to do this? Sharpe ratio? Stddev? I remember reading a post of your a few weeks back where you talked about measuring risk-adjusted return in some way but I'm a lemon with the serach facility.

If you have monthly, hourly, or whatever results from your own methodology, what's the best way to compare your actice results with those of a passive index investor or the like?

I'd hate to think I could have just replicated my last 5 years worth of results by leveraging the FTSE All-Share with an IMR of 50% with substantially less risk.

I've tested and (small scale) traded various strategies but I received 16% for my uni biometrics (stats) exam and will have to stop my scotch rambling.

[/ QUOTE ]

first off, what is a 16%? is that good or bad? (what is the class average?). seems like you should be killing that exam.

anyways, quantifying risk is art and skill. the new theories are not practical yet since the mathematics behind them are prohibitive and their estimates are so sensitive that any "banding" around reasonable outputs renders the estimates virtually unusuable (i.e. very very small changes in exponent estimates results in huge swings in the output distribution)

i'd approach your problem first obviously from a qualitative standpoint. what risks are you most susceptible to? i.e., what does your trading strategy do and how could it come undone? for here i'd look at the periods of the backtest where your strategy did the worst. what was going on in the strategy that caused the poor performance? what could, in the future, cause similar instances of underperformance.

then, calculate the daily/monthly/annualized st.dev for those periods and compare it to the st.dev figures of the surrounding better performing time periods. even more generally, calculate the daily, monthly, and annual rolling standard deviation of the backtest returns. just plot them and see how much each of them moves around (relative tot he time frame for your strategy...i don't know if you have a daily one so maybe that one isn't as important though you said you calculate risk on a daily basis)

if the data is really noisy (quite possible if not likely) smooth it out a bit and then zero in on the periods of biggest jumps in your "risk" measurement. then compare them to the poor performance periods and the surrounding periods to just try to grasp exactly what your strategy goes through during those time periods.

the goal here isn't to generate a specific #, but to first understand what that number could tell you. one example would be if your risk measurement in a shorter term trading strategy, when low, results in adding exposure. if, in the backtest, your periods of poorest performance occurr after long periods of low volatility followed by one brief large jump in risk it coudl be that you are allocating too much based on recent historic risk levels and not enough on the overall potential shocks to the system.

anyways, let me know if that ramble answered your question...

in terms of comparison, you basically said it. i'd just take your actual returns and subtract out whatever benchmark you'd want to compare it to and look at excess returns of your system (also, for all the above risk analysis you should be either looking at excess returns - subtractinga benchmark index- or excess returns - subtracting cash)

it may be the case that your system could be generating too many units of risk for every unit of excess return and a leveraged index appraoch would have done better on an absolute and risk adjusted basis. that would definitely suck but isn't out of the realm of possibility.

the excess returns (using say, the FTSE all share or 100 or whatever) of your system is the comparative statistic since you are comparing your excess returns to the benchmark's. you must obviously subtract cash from the benchmark's returns to compare it to your excess returns.

a quick sharpe calc here would suffice.

hope this helps. i'd be interested in hearing the results.

thanks,
Barron

[wrschultz]

Very good thread, even though some of it is a bit old.

I have to keep about $25,000 in deposit accounts at a major bank, and I'm now using Bank of America. They have a brokerage account deal where if you have 25k on deposit at BOA, you get 40 free trades per month.

Since I need this money on deposit anyway, this seems like a great deal. I know nothing about investing, and I'm wondering how I can use this, even with little to no knowledge.

Thanks

[DcifrThs]

[ QUOTE ]
Thanks for this interesting thread.

1) What are the most important valuation metrics that you use when determining buying a value stock on say the NYSE?

[/ QUOTE ]

i'm not an equity trader nor qualified to give individual stock advice. this is definitely a questino for desertcat.

[ QUOTE ]

2) Same question as #1 but a high risk high return stock?

[/ QUOTE ]

again, not qualified, sorry [img]/images/graemlins/frown.gif[/img]

[ QUOTE ]
Could you share your thought proccesses on how you buy stock? What information you use? What do you look for, etc, etc?

[/ QUOTE ]

i don't buy individual stocks so i can't answer that.

i can speak semi-intelligently about why stocks can or can't be purchased for excess returns based on fundamental analysis. i can also speak to some of the major fundamental drivers of the equity market so let me know if you want to rephrase your question or ask desertcat...

[ QUOTE ]


What does your sn mean?

[/ QUOTE ]

lol. i always get a kick when somebody asks this question [img]/images/graemlins/smile.gif[/img]

Barron

[DcifrThs]

[ QUOTE ]
Earlier this year, I moved a huge chunk of my 401(k) to an "inflation-protected securities" fund in a simplistic desire to escape equities, without knowing what the fund actually was. Yeah, not too smart.

So...having luckboxed into a nice return thus far, I'm wondering:

1) Why have inflation-protected securities done so well this year?

[/ QUOTE ]

inflation protection securities do well when inflation comes in above break even inflation or when inflation expectations increase (and investors are willing to accept a lower yield for the premium of protectino against inflation).

they also do well (just like their nominal counterparts) when real yields fall. one complicating factor here though is that if real yields fall too much, then inflation expectations may increase so TIPS can do proportionally better than their nominal counterparts during these times (and vice versa during late stage tightening)

further, since TIPS are govt issued, their credit quality is taken as risk free. thus, anytime huge jumps in volatility occur, all risk free assets benefit.

all of those have happened this year.

[ QUOTE ]


2) Which economic indicator(s) would portend a downturn in their value?

[/ QUOTE ]

inflation vs. break even inflation and the yield curve are the main determinants in their value. unfortunately, since the former can't be measured daily/weekly and the later is the realization of the former (either in fact or expectations) there is no indicator that can universally give you a "sell" signal.

in terms of economic indicators though, you could look at grwoth vs. trend in conjunction with inflationary pressures (industrial production, cap.util.) to try to tease out where the actual yield curve should be and where it is now. the "should be" of course is absent flows/momentum so you can't take that as given.

[ QUOTE ]
I tried reading the Wikipedia entry on TIPS and couldn't get past the first few sentences.

Thanks for all your great insights in this thread!

[/ QUOTE ]

glad to help..

i read the wiki on TIPS and it seems pretty clear while lacking any valuation/indicator stuff...but i hope i could add some clarity to it.

Barron

[DcifrThs]

[ QUOTE ]
Dcifrths,

Would you mind sharing some of the online resources that would be beneficial to noobs and Pros who want to further their understanding of investments, markets etc

[/ QUOTE ]

i'm not really sure other than my standard response in order of percieved value by me:

1) The Economist (print & online)
2) Financial Times (comment and analysis especially. read everything martin wolf ever wrote [img]/images/graemlins/smile.gif[/img]...the whole publication is good though so read as much as possible)
3) bloomberg.com- great for up to date market commentary and the "Economic Calander" is huge!!
4) wall street journal. this is more factual and X happened then Y happened type stuff vs. the FT which provides some more color as well as international flavor, which is important since globalization is such a huge factor in what is going on.

for instance, recently we saw GDP #s come out and US personal consumption drove a good portion of the increase. that consumption has implications for italy, japan, UK etc.

as it relates to japan, since the tightening credit conditions weren't a huge factor in the 3rd quarter in the US, it is expected that they will play a bigger role in the 4th quarter and beyond (reducing overall consumption). this hurts japanese economic expectations (since US is biggest export market and japanese GDP is disproportionately export driven since domestic consumption is still sluggish).

thus the BoJ left rates unchanged and will likely shelve rate increases for teh forseeable future.

so just read up constantly is the best i can say...
Barron

[kimchi]

Thanks for the detailed answer

[ QUOTE ]
first off, what is a 16%? is that good or bad? (what is the class average)

[/ QUOTE ]

Very bad (and bad). We were ecologists who had to grit our teeth through the mandatory stat classes. We all did so badly that the prof added 10% onto all our scores - one of my friends got 11% (ie-1%). You need to remember though that marking is much tougher in The UK. A first-class honours student from Oxford or Cambridge would probably get 70% and a pass would be above 35%.

[ QUOTE ]

what risks are you most susceptible to? i.e., what does your trading strategy do and how could it come undone? for here i'd look at the periods of the backtest where your strategy did the worst. what was going on in the strategy that caused the poor performance? what could, in the future, cause similar instances of underperformance.

[/ QUOTE ]

I think this is a useful excercise. My longer term strategy uses a smoothed relative strength ranking system I created. It needs a long look-back period (200 days) and be further smoothed by a short EMA to avoid regular transactions and the irritating whipsaws inherent in a trend-following strategy.

I can only trade long with this strategy (for tax & available investment vehicle reasons) but the breaks on the short side for stocks are more rapid and violent than the moves on the long side. This creates sudden drops which my system takes a while to catch up with and create a new signal.

1999 was the best year followed by 2000 which was the worst due to the system being invested in tech, Japan, Asia, and Latin Am stocks. I had to keep hold of most of these until around May or June of 2000 before a valid signal was issued to move into financials, health, bonds etc. creating large drawdowns.

Tthe worst (peak to trough) drawdown was during the summer sell-off of 2005 though as I was fully invested in Latin, general emerging, and asia stocks at the time. I have since reduced the size of each position and increased the number of positions to add some more diversity at the expense of a reduced return.

[ QUOTE ]
if, in the backtest, your periods of poorest performance occurr after long periods of low volatility followed by one brief large jump in risk it coudl be that you are allocating too much based on recent historic risk levels and not enough on the overall potential shocks to the system.

[/ QUOTE ]

I think this is a very important point. My other (swing) trading strategy defines risk as a 20-day period EMA of the ATR in the market I'm trading. I can then size my positions accordingly so if I had 5 positions open, each would expose my account to a similar amount of risk regardless of the volatility of the market. This means I could theoretically be exposed to the same risk whether I'm trading gilts or nanotech IPOs.

Sudden market shocks that blow through stops occur frequesntly enough so that I need stop-loss insurance which increases my trading costs quite significantly. I only trade EOD and so if George Bush chokes on a pretzel I don't want to wake up to a big fat red arrow on my screen.

[ QUOTE ]

hope this helps. i'd be interested in hearing the results.


[/ QUOTE ]

I showed you some results in a thread a while back. I have updated and modified reults which I will get to work on tonight and apply the suggestions you made.

Thanks for the useful info.

[DcifrThs]

kimchi,

glad i could help.

i would like to take a quick moment here and dig into something you learned through experience that it hink would be a useful lesson overall:

[ QUOTE ]
Tthe worst (peak to trough) drawdown was during the summer sell-off of 2005 though as I was fully invested in Latin, general emerging, and asia stocks at the time. I have since reduced the size of each position and increased the number of positions to add some more diversity at the expense of a reduced return.

[/ QUOTE ]

first off, working backwards, the system may not (probably not given above) take into account volatilities and correlations of markets. the reason i say this is because a huge portion of your system shouldn't be in markets that are both a) highly volatile, and b) highly correlated to each other and the global economy.

you haven't "sacrificed returns" either because those were returns you will pay for with huge peak to trough drawdowns on occasion given the volatility/correlation of those positions.

what kind of risk check do you have on the system? i.e. if X position => Y % of overall risk pie, cut X back to Y and realocate. this type of thing would definitely limit the drawdown you experienced and almost certainly increase your risk adjusted returns if you group and define your strategies well.

the main point i want to make with this tangent is that thinking solely in absolute return space can lead to issues liek the one you descibe. thinking more in risk space (which from previous posts i think you do most of the time..not sure about your system) is more useful and the allocation of risk should definitely trump the allocation fo capital.

lemme know yoru thoughts,
Barron