The Fire Dave unique interviews have proven extremely popular as more and more people in Ireland are searching for information about this emerging sub culture we call “Financial Independence and Early Retirement” or “FIRE”. I now know more than ever that you guys and girls very much look forward to when I release a new story and unique journey and the feedback you have given me has kept this going.
I went back to Jorge who I interviewed at the start of 2021 when he said:
“First off, I maxed out my pension contributions in late 2019 after a chat at work with a fellow expat. He opened my eyes to the fact that the Irish pension scheme is not a scam like the one in our home country.
Then I opened a brokerage account and started dollar-cost-averaging through the pandemic-fuelled volatility. My strategy is quite aggressive given how late I started: “
If you are similarly looking to open your own brokerage account check out my Quick Guide to DEGIRO for easy steps to open your own account.
Jorge kindly shared his Retirement Simulator the last time which has since been updated.
Interestingly, since we last spoke Jorge has added in the ability to “stress test” a market crash which makes this a very valuable tool for anyone on the FIRE journey and the level of detail is exceptional.
I also invited Jorge to be a guest speaker at our local Financial Independence Ireland, Limerick Meetup Group. Jorge will be presenting followed by Q&A and an informal discussion @8pm on 26.05.2021
RSVP here Financial Independence Ireland Online Meetup (Limerick) | Meetup
I am not going to spoil anymore of this FIRE interview so enjoy the latest updates and from me personally a huge thank you to Jorge for the effort involved and I urge you all to read over the first interview if you missed it and explore the Retirement Simulator whether you have opened a brokerage account yet or not…
How are you getting on with your FIRE journey now since the interview?
I have been following my plan of DCA (Dollar Cost Averaging) my entrance into the market, with only one significant tweak. I was too concentrated in the tech sector as part of an aggressive strategy.
The risk worried me though, and after much reading and running of scenarios, I decided to allocate a portion of my portfolio to an MSCI world index ETF. I didn’t want to rebalance, as that would cause a tax event, so I switched to buying that new ETF until it reaches its target allocation. I’m still on that journey, and will likely keep tweaking my allocation occasionally.
I don’t expect to do this frequently though, in part because I hope to settle on a strategy that I feel comfortable with, and also because I stopped following the market news as closely as I was. After reading The Psychology of Money by Morgan Housel I realised that, as an investor, I’m not the target audience of the financial news cycle, aimed at traders. I need the mental space to stay focused on my job, on which my retirement plan depends, so I filtered all that out.
I keep in touch with the FIRE community though, as it’s more about the long term and there’s always interesting insights, and I try to keep tabs on the macro picture. On that front, I enjoy Paul Donovan’s daily email. It’s calm, short, and witty.
Have you developed and utilised the retirement simulator further since we spoke?
One of the ideas I had in the backlog was to run my scenario with actual historic market data instead of using an artificial constant growth, but I couldn’t figure out how to make that generic enough since the market performance depends on what you invest in.
I was tinkering at work with a way to improve software delivery forecasts through the use of Monte Carlo simulations, which is an estimation method that runs a calculation many times using random or historic data and counts how many of those runs meet some success criteria (delivery by a certain date in this case). Long story short, I confirmed the well known fact that software estimation is a hard problem and can’t be done reliably. But I realised that this technique can be applied to the financial simulator.
The random walk of the stock market can be simulated with a statistical distribution defined by the average return and volatility, so I added the ability to define those parameters for each investment type. The simulator then runs through your scenario thousands of times, each one following a different random market walk, and counts how many of them reach the target final age (a grim but necessary parameter) before running out of money. In other words, given the market conditions you set up, how likely is your plan to succeed.
I also added the ability to simulate market crashes, because even though analysts’ constant predictions of imminent bubble bursts never seem to materialize (last March’s pandemic-fuelled dip notwithstanding), eventually crashes do happen. They can last years and disrupt the best laid out plans. I wanted to know the impact of a crash while I was entering the market or right after, right before or after retirement, or how survivable a crash every ten years would be.
Now, if you look up the historic volatility of, say, the S&P 500, it already includes the 2008 financial crisis, the 2001 dot com crash and all the crashes before that. So I was adding even more crashes on top of the ones already embedded in the volatility. I mean, if I’m going to stress test my plan, I might as well go all the way and simulate the economy of my home country. Suffice to say, my plan would not survive Ireland’s tax system and cost of living if it adopted my home country’s fondness for wild economic crises.
What are the future plans for the simulator?
I already got a lot out of it. It helped me understand what I need to do in order to avoid starving after retirement, and it gave me peace of mind once I realized that I can avoid that.
I used it to complement and validate the recommendations of my financial advisor.
I’ve also used it in a talk I gave titled “what I would tell my younger self” to illustrate the importance of not leaving your savings sitting in the bank, so it even works as an educational tool.
But there’s a list of known issues (you can find them near the bottom of the user guide). Most of them are limitations that come from my personal situation, since I built it for myself first and foremost. I have received help from people who pointed out issues, as witnessed by the change log at the end of the user guide, and a fellow software engineer contributed an enhancement (the source code is on github).
I would love for more people to contribute and improve it, make it more general and useful so others can get out of it as much as I did. Young software engineer: if you’re looking for a side-project to include in your LinkedIn profile, here’s your chance! 🙂